* [PATCH 01/55] arm/arm64: KVM: Add a protection parameter to create_hyp_mappings
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 02/55] arm64: Add PTE_HYP_XN page table flag Marc Zyngier
` (54 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Currently, create_hyp_mappings applies a "one size fits all" page
protection (PAGE_HYP). As we're heading towards separate protections
for different sections, let's make this protection a parameter, and
let the callers pass their prefered protection (PAGE_HYP for everyone
for the time being).
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_mmu.h | 2 +-
arch/arm/kvm/arm.c | 13 +++++++------
arch/arm/kvm/mmu.c | 5 +++--
arch/arm64/include/asm/kvm_mmu.h | 2 +-
4 files changed, 12 insertions(+), 10 deletions(-)
diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h
index f9a6506..6cb4d4d 100644
--- a/arch/arm/include/asm/kvm_mmu.h
+++ b/arch/arm/include/asm/kvm_mmu.h
@@ -49,7 +49,7 @@
#include <asm/pgalloc.h>
#include <asm/stage2_pgtable.h>
-int create_hyp_mappings(void *from, void *to);
+int create_hyp_mappings(void *from, void *to, pgprot_t prot);
int create_hyp_io_mappings(void *from, void *to, phys_addr_t);
void free_boot_hyp_pgd(void);
void free_hyp_pgds(void);
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index f20ca84..45dd6df 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -122,7 +122,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
if (ret)
goto out_fail_alloc;
- ret = create_hyp_mappings(kvm, kvm + 1);
+ ret = create_hyp_mappings(kvm, kvm + 1, PAGE_HYP);
if (ret)
goto out_free_stage2_pgd;
@@ -239,7 +239,7 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
if (err)
goto free_vcpu;
- err = create_hyp_mappings(vcpu, vcpu + 1);
+ err = create_hyp_mappings(vcpu, vcpu + 1, PAGE_HYP);
if (err)
goto vcpu_uninit;
@@ -1293,14 +1293,14 @@ static int init_hyp_mode(void)
* Map the Hyp-code called directly from the host
*/
err = create_hyp_mappings(kvm_ksym_ref(__hyp_text_start),
- kvm_ksym_ref(__hyp_text_end));
+ kvm_ksym_ref(__hyp_text_end), PAGE_HYP);
if (err) {
kvm_err("Cannot map world-switch code\n");
goto out_err;
}
err = create_hyp_mappings(kvm_ksym_ref(__start_rodata),
- kvm_ksym_ref(__end_rodata));
+ kvm_ksym_ref(__end_rodata), PAGE_HYP);
if (err) {
kvm_err("Cannot map rodata section\n");
goto out_err;
@@ -1311,7 +1311,8 @@ static int init_hyp_mode(void)
*/
for_each_possible_cpu(cpu) {
char *stack_page = (char *)per_cpu(kvm_arm_hyp_stack_page, cpu);
- err = create_hyp_mappings(stack_page, stack_page + PAGE_SIZE);
+ err = create_hyp_mappings(stack_page, stack_page + PAGE_SIZE,
+ PAGE_HYP);
if (err) {
kvm_err("Cannot map hyp stack\n");
@@ -1323,7 +1324,7 @@ static int init_hyp_mode(void)
kvm_cpu_context_t *cpu_ctxt;
cpu_ctxt = per_cpu_ptr(kvm_host_cpu_state, cpu);
- err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1);
+ err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1, PAGE_HYP);
if (err) {
kvm_err("Cannot map host CPU state: %d\n", err);
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 45c43ae..49cb5cc 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -679,12 +679,13 @@ static phys_addr_t kvm_kaddr_to_phys(void *kaddr)
* create_hyp_mappings - duplicate a kernel virtual address range in Hyp mode
* @from: The virtual kernel start address of the range
* @to: The virtual kernel end address of the range (exclusive)
+ * @prot: The protection to be applied to this range
*
* The same virtual address as the kernel virtual address is also used
* in Hyp-mode mapping (modulo HYP_PAGE_OFFSET) to the same underlying
* physical pages.
*/
-int create_hyp_mappings(void *from, void *to)
+int create_hyp_mappings(void *from, void *to, pgprot_t prot)
{
phys_addr_t phys_addr;
unsigned long virt_addr;
@@ -704,7 +705,7 @@ int create_hyp_mappings(void *from, void *to)
err = __create_hyp_mappings(hyp_pgd, virt_addr,
virt_addr + PAGE_SIZE,
__phys_to_pfn(phys_addr),
- PAGE_HYP);
+ prot);
if (err)
return err;
}
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index f05ac27..fdfbddb 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -81,7 +81,7 @@ alternative_endif
#include <asm/stage2_pgtable.h>
-int create_hyp_mappings(void *from, void *to);
+int create_hyp_mappings(void *from, void *to, pgprot_t prot);
int create_hyp_io_mappings(void *from, void *to, phys_addr_t);
void free_boot_hyp_pgd(void);
void free_hyp_pgds(void);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 02/55] arm64: Add PTE_HYP_XN page table flag
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
2016-07-22 17:28 ` [PATCH 01/55] arm/arm64: KVM: Add a protection parameter to create_hyp_mappings Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 03/55] arm/arm64: KVM: Enforce HYP read-only mapping of the kernel's rodata section Marc Zyngier
` (53 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
EL2 page tables can be configured to deny code from being
executed, which is done by setting bit 54 in the page descriptor.
It is the same bit as PTE_UXN, but the "USER" reference felt odd
in the hypervisor code.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/pgtable-hwdef.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/arch/arm64/include/asm/pgtable-hwdef.h b/arch/arm64/include/asm/pgtable-hwdef.h
index 2813748..c3ae239 100644
--- a/arch/arm64/include/asm/pgtable-hwdef.h
+++ b/arch/arm64/include/asm/pgtable-hwdef.h
@@ -164,6 +164,7 @@
#define PTE_CONT (_AT(pteval_t, 1) << 52) /* Contiguous range */
#define PTE_PXN (_AT(pteval_t, 1) << 53) /* Privileged XN */
#define PTE_UXN (_AT(pteval_t, 1) << 54) /* User XN */
+#define PTE_HYP_XN (_AT(pteval_t, 1) << 54) /* HYP XN */
/*
* AttrIndx[2:0] encoding (mapping attributes defined in the MAIR* registers).
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 03/55] arm/arm64: KVM: Enforce HYP read-only mapping of the kernel's rodata section
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
2016-07-22 17:28 ` [PATCH 01/55] arm/arm64: KVM: Add a protection parameter to create_hyp_mappings Marc Zyngier
2016-07-22 17:28 ` [PATCH 02/55] arm64: Add PTE_HYP_XN page table flag Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 04/55] arm/arm64: KVM: Map the HYP text as read-only Marc Zyngier
` (52 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
In order to be able to use C code in HYP, we're now mapping the kernel's
rodata in HYP. It works absolutely fine, except that we're mapping it RWX,
which is not what it should be.
Add a new HYP_PAGE_RO protection, and pass it as the protection flags
when mapping the rodata section.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/pgtable.h | 1 +
arch/arm/kvm/arm.c | 2 +-
arch/arm64/include/asm/pgtable-prot.h | 1 +
3 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h
index 348caab..f332087 100644
--- a/arch/arm/include/asm/pgtable.h
+++ b/arch/arm/include/asm/pgtable.h
@@ -98,6 +98,7 @@ extern pgprot_t pgprot_s2_device;
#define PAGE_KERNEL _MOD_PROT(pgprot_kernel, L_PTE_XN)
#define PAGE_KERNEL_EXEC pgprot_kernel
#define PAGE_HYP _MOD_PROT(pgprot_kernel, L_PTE_HYP)
+#define PAGE_HYP_RO _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY | L_PTE_XN)
#define PAGE_HYP_DEVICE _MOD_PROT(pgprot_hyp_device, L_PTE_HYP)
#define PAGE_S2 _MOD_PROT(pgprot_s2, L_PTE_S2_RDONLY)
#define PAGE_S2_DEVICE _MOD_PROT(pgprot_s2_device, L_PTE_S2_RDONLY)
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 45dd6df..b308976 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -1300,7 +1300,7 @@ static int init_hyp_mode(void)
}
err = create_hyp_mappings(kvm_ksym_ref(__start_rodata),
- kvm_ksym_ref(__end_rodata), PAGE_HYP);
+ kvm_ksym_ref(__end_rodata), PAGE_HYP_RO);
if (err) {
kvm_err("Cannot map rodata section\n");
goto out_err;
diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h
index 29fcb33..88db58c 100644
--- a/arch/arm64/include/asm/pgtable-prot.h
+++ b/arch/arm64/include/asm/pgtable-prot.h
@@ -56,6 +56,7 @@
#define PAGE_KERNEL_EXEC_CONT __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT)
#define PAGE_HYP __pgprot(_PAGE_DEFAULT | PTE_HYP)
+#define PAGE_HYP_RO __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY | PTE_HYP_XN)
#define PAGE_HYP_DEVICE __pgprot(PROT_DEVICE_nGnRE | PTE_HYP)
#define PAGE_S2 __pgprot(PROT_DEFAULT | PTE_S2_MEMATTR(MT_S2_NORMAL) | PTE_S2_RDONLY)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 04/55] arm/arm64: KVM: Map the HYP text as read-only
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (2 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 03/55] arm/arm64: KVM: Enforce HYP read-only mapping of the kernel's rodata section Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 05/55] arm/arm64: KVM: Make default HYP mappings non-excutable Marc Zyngier
` (51 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
There should be no reason for mapping the HYP text read/write.
As such, let's have a new set of flags (PAGE_HYP_EXEC) that allows
execution, but makes the page as read-only, and update the two call
sites that deal with mapping code.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/pgtable.h | 1 +
arch/arm/kvm/arm.c | 2 +-
arch/arm/kvm/mmu.c | 6 +++---
arch/arm64/include/asm/pgtable-prot.h | 1 +
4 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h
index f332087..7487bf9 100644
--- a/arch/arm/include/asm/pgtable.h
+++ b/arch/arm/include/asm/pgtable.h
@@ -98,6 +98,7 @@ extern pgprot_t pgprot_s2_device;
#define PAGE_KERNEL _MOD_PROT(pgprot_kernel, L_PTE_XN)
#define PAGE_KERNEL_EXEC pgprot_kernel
#define PAGE_HYP _MOD_PROT(pgprot_kernel, L_PTE_HYP)
+#define PAGE_HYP_EXEC _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY)
#define PAGE_HYP_RO _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY | L_PTE_XN)
#define PAGE_HYP_DEVICE _MOD_PROT(pgprot_hyp_device, L_PTE_HYP)
#define PAGE_S2 _MOD_PROT(pgprot_s2, L_PTE_S2_RDONLY)
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index b308976..c74483f 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -1293,7 +1293,7 @@ static int init_hyp_mode(void)
* Map the Hyp-code called directly from the host
*/
err = create_hyp_mappings(kvm_ksym_ref(__hyp_text_start),
- kvm_ksym_ref(__hyp_text_end), PAGE_HYP);
+ kvm_ksym_ref(__hyp_text_end), PAGE_HYP_EXEC);
if (err) {
kvm_err("Cannot map world-switch code\n");
goto out_err;
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 49cb5cc..679608f 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -1733,7 +1733,7 @@ int kvm_mmu_init(void)
err = __create_hyp_mappings(boot_hyp_pgd,
hyp_idmap_start, hyp_idmap_end,
__phys_to_pfn(hyp_idmap_start),
- PAGE_HYP);
+ PAGE_HYP_EXEC);
if (err) {
kvm_err("Failed to idmap %lx-%lx\n",
@@ -1756,7 +1756,7 @@ int kvm_mmu_init(void)
err = __create_hyp_mappings(boot_hyp_pgd,
TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE,
__phys_to_pfn(hyp_idmap_start),
- PAGE_HYP);
+ PAGE_HYP_EXEC);
if (err) {
kvm_err("Failed to map trampoline @%lx into boot HYP pgd\n",
TRAMPOLINE_VA);
@@ -1767,7 +1767,7 @@ int kvm_mmu_init(void)
err = __create_hyp_mappings(hyp_pgd,
TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE,
__phys_to_pfn(hyp_idmap_start),
- PAGE_HYP);
+ PAGE_HYP_EXEC);
if (err) {
kvm_err("Failed to map trampoline @%lx into runtime HYP pgd\n",
TRAMPOLINE_VA);
diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h
index 88db58c..3802048 100644
--- a/arch/arm64/include/asm/pgtable-prot.h
+++ b/arch/arm64/include/asm/pgtable-prot.h
@@ -56,6 +56,7 @@
#define PAGE_KERNEL_EXEC_CONT __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT)
#define PAGE_HYP __pgprot(_PAGE_DEFAULT | PTE_HYP)
+#define PAGE_HYP_EXEC __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY)
#define PAGE_HYP_RO __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY | PTE_HYP_XN)
#define PAGE_HYP_DEVICE __pgprot(PROT_DEVICE_nGnRE | PTE_HYP)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 05/55] arm/arm64: KVM: Make default HYP mappings non-excutable
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (3 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 04/55] arm/arm64: KVM: Map the HYP text as read-only Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 06/55] KVM: arm/arm64: The GIC is dead, long live the GIC Marc Zyngier
` (50 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Structures that can be generally written to don't have any requirement
to be executable (quite the opposite). This includes the kvm and vcpu
structures, as well as the stacks.
Let's change the default to incorporate the XN flag.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/pgtable.h | 2 +-
arch/arm64/include/asm/pgtable-prot.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h
index 7487bf9..e0d76ba 100644
--- a/arch/arm/include/asm/pgtable.h
+++ b/arch/arm/include/asm/pgtable.h
@@ -97,7 +97,7 @@ extern pgprot_t pgprot_s2_device;
#define PAGE_READONLY_EXEC _MOD_PROT(pgprot_user, L_PTE_USER | L_PTE_RDONLY)
#define PAGE_KERNEL _MOD_PROT(pgprot_kernel, L_PTE_XN)
#define PAGE_KERNEL_EXEC pgprot_kernel
-#define PAGE_HYP _MOD_PROT(pgprot_kernel, L_PTE_HYP)
+#define PAGE_HYP _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_XN)
#define PAGE_HYP_EXEC _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY)
#define PAGE_HYP_RO _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY | L_PTE_XN)
#define PAGE_HYP_DEVICE _MOD_PROT(pgprot_hyp_device, L_PTE_HYP)
diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h
index 3802048..39f5252 100644
--- a/arch/arm64/include/asm/pgtable-prot.h
+++ b/arch/arm64/include/asm/pgtable-prot.h
@@ -55,7 +55,7 @@
#define PAGE_KERNEL_EXEC __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE)
#define PAGE_KERNEL_EXEC_CONT __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT)
-#define PAGE_HYP __pgprot(_PAGE_DEFAULT | PTE_HYP)
+#define PAGE_HYP __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_HYP_XN)
#define PAGE_HYP_EXEC __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY)
#define PAGE_HYP_RO __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY | PTE_HYP_XN)
#define PAGE_HYP_DEVICE __pgprot(PROT_DEVICE_nGnRE | PTE_HYP)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 06/55] KVM: arm/arm64: The GIC is dead, long live the GIC
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (4 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 05/55] arm/arm64: KVM: Make default HYP mappings non-excutable Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 07/55] arm64: KVM: Merged page tables documentation Marc Zyngier
` (49 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
I don't think any single piece of the KVM/ARM code ever generated
as much hatred as the GIC emulation.
It was written by someone who had zero experience in modeling
hardware (me), was riddled with design flaws, should have been
scrapped and rewritten from scratch long before having a remote
chance of reaching mainline, and yet we supported it for a good
three years. No need to mention the names of those who suffered,
the git log is singing their praises.
Thankfully, we now have a much more maintainable implementation,
and we can safely put the grumpy old GIC to rest.
Fellow hackers, please raise your glass in memory of the GIC:
The GIC is dead, long live the GIC!
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/kvm/Kconfig | 7 -
arch/arm/kvm/Makefile | 6 -
arch/arm64/kvm/Kconfig | 7 -
arch/arm64/kvm/Makefile | 8 -
include/kvm/{vgic/vgic.h => arm_vgic.h} | 6 +-
virt/kvm/arm/hyp/vgic-v2-sr.c | 15 +-
virt/kvm/arm/vgic-v2-emul.c | 856 -----------
virt/kvm/arm/vgic-v2.c | 274 ----
virt/kvm/arm/vgic-v3-emul.c | 1074 --------------
virt/kvm/arm/vgic-v3.c | 279 ----
virt/kvm/arm/vgic.c | 2440 -------------------------------
virt/kvm/arm/vgic.h | 140 --
12 files changed, 7 insertions(+), 5105 deletions(-)
rename include/kvm/{vgic/vgic.h => arm_vgic.h} (98%)
delete mode 100644 virt/kvm/arm/vgic-v2-emul.c
delete mode 100644 virt/kvm/arm/vgic-v2.c
delete mode 100644 virt/kvm/arm/vgic-v3-emul.c
delete mode 100644 virt/kvm/arm/vgic-v3.c
delete mode 100644 virt/kvm/arm/vgic.c
delete mode 100644 virt/kvm/arm/vgic.h
diff --git a/arch/arm/kvm/Kconfig b/arch/arm/kvm/Kconfig
index 02abfff..95a0005 100644
--- a/arch/arm/kvm/Kconfig
+++ b/arch/arm/kvm/Kconfig
@@ -46,13 +46,6 @@ config KVM_ARM_HOST
---help---
Provides host support for ARM processors.
-config KVM_NEW_VGIC
- bool "New VGIC implementation"
- depends on KVM
- default y
- ---help---
- uses the new VGIC implementation
-
source drivers/vhost/Kconfig
endif # VIRTUALIZATION
diff --git a/arch/arm/kvm/Makefile b/arch/arm/kvm/Makefile
index a596b58..5e28df8 100644
--- a/arch/arm/kvm/Makefile
+++ b/arch/arm/kvm/Makefile
@@ -22,7 +22,6 @@ obj-y += kvm-arm.o init.o interrupts.o
obj-y += arm.o handle_exit.o guest.o mmu.o emulate.o reset.o
obj-y += coproc.o coproc_a15.o coproc_a7.o mmio.o psci.o perf.o
-ifeq ($(CONFIG_KVM_NEW_VGIC),y)
obj-y += $(KVM)/arm/vgic/vgic.o
obj-y += $(KVM)/arm/vgic/vgic-init.o
obj-y += $(KVM)/arm/vgic/vgic-irqfd.o
@@ -30,9 +29,4 @@ obj-y += $(KVM)/arm/vgic/vgic-v2.o
obj-y += $(KVM)/arm/vgic/vgic-mmio.o
obj-y += $(KVM)/arm/vgic/vgic-mmio-v2.o
obj-y += $(KVM)/arm/vgic/vgic-kvm-device.o
-else
-obj-y += $(KVM)/arm/vgic.o
-obj-y += $(KVM)/arm/vgic-v2.o
-obj-y += $(KVM)/arm/vgic-v2-emul.o
-endif
obj-y += $(KVM)/arm/arch_timer.o
diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
index c4f26ef..aa2e34e 100644
--- a/arch/arm64/kvm/Kconfig
+++ b/arch/arm64/kvm/Kconfig
@@ -54,13 +54,6 @@ config KVM_ARM_PMU
Adds support for a virtual Performance Monitoring Unit (PMU) in
virtual machines.
-config KVM_NEW_VGIC
- bool "New VGIC implementation"
- depends on KVM
- default y
- ---help---
- uses the new VGIC implementation
-
source drivers/vhost/Kconfig
endif # VIRTUALIZATION
diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
index a7a958c..f00b2cd 100644
--- a/arch/arm64/kvm/Makefile
+++ b/arch/arm64/kvm/Makefile
@@ -20,7 +20,6 @@ kvm-$(CONFIG_KVM_ARM_HOST) += emulate.o inject_fault.o regmap.o
kvm-$(CONFIG_KVM_ARM_HOST) += hyp.o hyp-init.o handle_exit.o
kvm-$(CONFIG_KVM_ARM_HOST) += guest.o debug.o reset.o sys_regs.o sys_regs_generic_v8.o
-ifeq ($(CONFIG_KVM_NEW_VGIC),y)
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-init.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-irqfd.o
@@ -30,12 +29,5 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v2.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v3.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-kvm-device.o
-else
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v2.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v2-emul.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v3.o
-kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v3-emul.o
-endif
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
diff --git a/include/kvm/vgic/vgic.h b/include/kvm/arm_vgic.h
similarity index 98%
rename from include/kvm/vgic/vgic.h
rename to include/kvm/arm_vgic.h
index 3fbd175..1264037 100644
--- a/include/kvm/vgic/vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -13,8 +13,8 @@
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
-#ifndef __ASM_ARM_KVM_VGIC_VGIC_H
-#define __ASM_ARM_KVM_VGIC_VGIC_H
+#ifndef __KVM_ARM_VGIC_H
+#define __KVM_ARM_VGIC_H
#include <linux/kernel.h>
#include <linux/kvm.h>
@@ -243,4 +243,4 @@ static inline int kvm_vgic_get_max_vcpus(void)
return kvm_vgic_global_state.max_gic_vcpus;
}
-#endif /* __ASM_ARM_KVM_VGIC_VGIC_H */
+#endif /* __KVM_ARM_VGIC_H */
diff --git a/virt/kvm/arm/hyp/vgic-v2-sr.c b/virt/kvm/arm/hyp/vgic-v2-sr.c
index 3a3a699..7cffd93 100644
--- a/virt/kvm/arm/hyp/vgic-v2-sr.c
+++ b/virt/kvm/arm/hyp/vgic-v2-sr.c
@@ -21,18 +21,11 @@
#include <asm/kvm_hyp.h>
-#ifdef CONFIG_KVM_NEW_VGIC
-extern struct vgic_global kvm_vgic_global_state;
-#define vgic_v2_params kvm_vgic_global_state
-#else
-extern struct vgic_params vgic_v2_params;
-#endif
-
static void __hyp_text save_maint_int_state(struct kvm_vcpu *vcpu,
void __iomem *base)
{
struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
- int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+ int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
u32 eisr0, eisr1;
int i;
bool expect_mi;
@@ -74,7 +67,7 @@ static void __hyp_text save_maint_int_state(struct kvm_vcpu *vcpu,
static void __hyp_text save_elrsr(struct kvm_vcpu *vcpu, void __iomem *base)
{
struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
- int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+ int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
u32 elrsr0, elrsr1;
elrsr0 = readl_relaxed(base + GICH_ELRSR0);
@@ -93,7 +86,7 @@ static void __hyp_text save_elrsr(struct kvm_vcpu *vcpu, void __iomem *base)
static void __hyp_text save_lrs(struct kvm_vcpu *vcpu, void __iomem *base)
{
struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
- int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+ int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
int i;
for (i = 0; i < nr_lr; i++) {
@@ -147,7 +140,7 @@ void __hyp_text __vgic_v2_restore_state(struct kvm_vcpu *vcpu)
struct vgic_v2_cpu_if *cpu_if = &vcpu->arch.vgic_cpu.vgic_v2;
struct vgic_dist *vgic = &kvm->arch.vgic;
void __iomem *base = kern_hyp_va(vgic->vctrl_base);
- int nr_lr = (kern_hyp_va(&vgic_v2_params))->nr_lr;
+ int nr_lr = (kern_hyp_va(&kvm_vgic_global_state))->nr_lr;
int i;
u64 live_lrs = 0;
diff --git a/virt/kvm/arm/vgic-v2-emul.c b/virt/kvm/arm/vgic-v2-emul.c
deleted file mode 100644
index 1b0bee0..0000000
diff --git a/virt/kvm/arm/vgic-v2.c b/virt/kvm/arm/vgic-v2.c
deleted file mode 100644
index 334cd7a..0000000
diff --git a/virt/kvm/arm/vgic-v3-emul.c b/virt/kvm/arm/vgic-v3-emul.c
deleted file mode 100644
index e661e7fb..0000000
diff --git a/virt/kvm/arm/vgic-v3.c b/virt/kvm/arm/vgic-v3.c
deleted file mode 100644
index 75b02fa..0000000
diff --git a/virt/kvm/arm/vgic.c b/virt/kvm/arm/vgic.c
deleted file mode 100644
index c3bfbb9..0000000
diff --git a/virt/kvm/arm/vgic.h b/virt/kvm/arm/vgic.h
deleted file mode 100644
index 0df74cb..0000000
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 07/55] arm64: KVM: Merged page tables documentation
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (5 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 06/55] KVM: arm/arm64: The GIC is dead, long live the GIC Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 08/55] arm64: KVM: Always reference __hyp_panic_string via its kernel VA Marc Zyngier
` (48 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Since dealing with VA ranges tends to hurt my brain badly, let's
start with a bit of documentation that will hopefully help
understanding what comes next...
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/kvm_mmu.h | 40 +++++++++++++++++++++++++++++++++++++---
1 file changed, 37 insertions(+), 3 deletions(-)
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index fdfbddb..6149dfc 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -29,10 +29,44 @@
*
* Instead, give the HYP mode its own VA region at a fixed offset from
* the kernel by just masking the top bits (which are all ones for a
- * kernel address).
+ * kernel address). We need to find out how many bits to mask.
*
- * ARMv8.1 (using VHE) does have a TTBR1_EL2, and doesn't use these
- * macros (the entire kernel runs at EL2).
+ * We want to build a set of page tables that cover both parts of the
+ * idmap (the trampoline page used to initialize EL2), and our normal
+ * runtime VA space, at the same time.
+ *
+ * Given that the kernel uses VA_BITS for its entire address space,
+ * and that half of that space (VA_BITS - 1) is used for the linear
+ * mapping, we can also limit the EL2 space to (VA_BITS - 1).
+ *
+ * The main question is "Within the VA_BITS space, does EL2 use the
+ * top or the bottom half of that space to shadow the kernel's linear
+ * mapping?". As we need to idmap the trampoline page, this is
+ * determined by the range in which this page lives.
+ *
+ * If the page is in the bottom half, we have to use the top half. If
+ * the page is in the top half, we have to use the bottom half:
+ *
+ * T = __virt_to_phys(__hyp_idmap_text_start)
+ * if (T & BIT(VA_BITS - 1))
+ * HYP_VA_MIN = 0 //idmap in upper half
+ * else
+ * HYP_VA_MIN = 1 << (VA_BITS - 1)
+ * HYP_VA_MAX = HYP_VA_MIN + (1 << (VA_BITS - 1)) - 1
+ *
+ * This of course assumes that the trampoline page exists within the
+ * VA_BITS range. If it doesn't, then it means we're in the odd case
+ * where the kernel idmap (as well as HYP) uses more levels than the
+ * kernel runtime page tables (as seen when the kernel is configured
+ * for 4k pages, 39bits VA, and yet memory lives just above that
+ * limit, forcing the idmap to use 4 levels of page tables while the
+ * kernel itself only uses 3). In this particular case, it doesn't
+ * matter which side of VA_BITS we use, as we're guaranteed not to
+ * conflict with anything.
+ *
+ * When using VHE, there are no separate hyp mappings and all KVM
+ * functionality is already mapped as part of the main kernel
+ * mappings, and none of this applies in that case.
*/
#define HYP_PAGE_OFFSET_SHIFT VA_BITS
#define HYP_PAGE_OFFSET_MASK ((UL(1) << HYP_PAGE_OFFSET_SHIFT) - 1)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 08/55] arm64: KVM: Always reference __hyp_panic_string via its kernel VA
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (6 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 07/55] arm64: KVM: Merged page tables documentation Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 09/55] arm/arm64: KVM: Remove hyp_kern_va helper Marc Zyngier
` (47 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
__hyp_panic_string is passed via the HYP panic code to the panic
function, and is being "upgraded" to a kernel address, as it is
referenced by the HYP code (in a PC-relative way).
This is a bit silly, and we'd be better off obtaining the kernel
address and not mess with it at all. This patch implements this
with a tiny bit of asm glue, by forcing the string pointer to be
read from the literal pool.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/kvm/hyp/switch.c | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/kvm/hyp/switch.c b/arch/arm64/kvm/hyp/switch.c
index 437cfad..81f21a2 100644
--- a/arch/arm64/kvm/hyp/switch.c
+++ b/arch/arm64/kvm/hyp/switch.c
@@ -299,9 +299,16 @@ static const char __hyp_panic_string[] = "HYP panic:\nPS:%08llx PC:%016llx ESR:%
static void __hyp_text __hyp_call_panic_nvhe(u64 spsr, u64 elr, u64 par)
{
- unsigned long str_va = (unsigned long)__hyp_panic_string;
+ unsigned long str_va;
- __hyp_do_panic(hyp_kern_va(str_va),
+ /*
+ * Force the panic string to be loaded from the literal pool,
+ * making sure it is a kernel address and not a PC-relative
+ * reference.
+ */
+ asm volatile("ldr %0, =__hyp_panic_string" : "=r" (str_va));
+
+ __hyp_do_panic(str_va,
spsr, elr,
read_sysreg(esr_el2), read_sysreg_el2(far),
read_sysreg(hpfar_el2), par,
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 09/55] arm/arm64: KVM: Remove hyp_kern_va helper
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (7 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 08/55] arm64: KVM: Always reference __hyp_panic_string via its kernel VA Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 10/55] arm64: KVM: Kill HYP_PAGE_OFFSET Marc Zyngier
` (46 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
hyp_kern_va is now completely unused, so let's remove it entirely.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_hyp.h | 1 -
arch/arm64/include/asm/kvm_hyp.h | 12 ------------
2 files changed, 13 deletions(-)
diff --git a/arch/arm/include/asm/kvm_hyp.h b/arch/arm/include/asm/kvm_hyp.h
index f0e8607..e38fce2 100644
--- a/arch/arm/include/asm/kvm_hyp.h
+++ b/arch/arm/include/asm/kvm_hyp.h
@@ -26,7 +26,6 @@
#define __hyp_text __section(.hyp.text) notrace
#define kern_hyp_va(v) (v)
-#define hyp_kern_va(v) (v)
#define __ACCESS_CP15(CRn, Op1, CRm, Op2) \
"mrc", "mcr", __stringify(p15, Op1, %0, CRn, CRm, Op2), u32
diff --git a/arch/arm64/include/asm/kvm_hyp.h b/arch/arm64/include/asm/kvm_hyp.h
index 44eaff7..1d81f9a 100644
--- a/arch/arm64/include/asm/kvm_hyp.h
+++ b/arch/arm64/include/asm/kvm_hyp.h
@@ -36,18 +36,6 @@ static inline unsigned long __kern_hyp_va(unsigned long v)
#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v)))
-static inline unsigned long __hyp_kern_va(unsigned long v)
-{
- u64 offset = PAGE_OFFSET - HYP_PAGE_OFFSET;
- asm volatile(ALTERNATIVE("add %0, %0, %1",
- "nop",
- ARM64_HAS_VIRT_HOST_EXTN)
- : "+r" (v) : "r" (offset));
- return v;
-}
-
-#define hyp_kern_va(v) (typeof(v))(__hyp_kern_va((unsigned long)(v)))
-
#define read_sysreg_elx(r,nvh,vh) \
({ \
u64 reg; \
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 10/55] arm64: KVM: Kill HYP_PAGE_OFFSET
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (8 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 09/55] arm/arm64: KVM: Remove hyp_kern_va helper Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 11/55] arm64: Add ARM64_HYP_OFFSET_LOW capability Marc Zyngier
` (45 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
HYP_PAGE_OFFSET is not massively useful. And the way we use it
in KERN_HYP_VA is inconsistent with the equivalent operation in
EL2, where we use a mask instead.
Let's replace the uses of HYP_PAGE_OFFSET with HYP_PAGE_OFFSET_MASK,
and get rid of the pointless macro.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/kvm_mmu.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 6149dfc..2f1e1ae 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -70,7 +70,6 @@
*/
#define HYP_PAGE_OFFSET_SHIFT VA_BITS
#define HYP_PAGE_OFFSET_MASK ((UL(1) << HYP_PAGE_OFFSET_SHIFT) - 1)
-#define HYP_PAGE_OFFSET (PAGE_OFFSET & HYP_PAGE_OFFSET_MASK)
/*
* Our virtual mapping for the idmap-ed MMU-enable code. Must be
@@ -104,7 +103,7 @@ alternative_endif
#include <asm/mmu_context.h>
#include <asm/pgtable.h>
-#define KERN_TO_HYP(kva) ((unsigned long)kva - PAGE_OFFSET + HYP_PAGE_OFFSET)
+#define KERN_TO_HYP(kva) ((unsigned long)kva & HYP_PAGE_OFFSET_MASK)
/*
* We currently only support a 40bit IPA.
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 11/55] arm64: Add ARM64_HYP_OFFSET_LOW capability
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (9 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 10/55] arm64: KVM: Kill HYP_PAGE_OFFSET Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 12/55] arm64: KVM: Define HYP offset masks Marc Zyngier
` (44 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
As we need to indicate to the rest of the kernel which region of
the HYP VA space is safe to use, add a capability that will
indicate that KVM should use the [VA_BITS-2:0] range.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/cpufeature.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/include/asm/cpufeature.h b/arch/arm64/include/asm/cpufeature.h
index 224efe7..d40edbb 100644
--- a/arch/arm64/include/asm/cpufeature.h
+++ b/arch/arm64/include/asm/cpufeature.h
@@ -36,8 +36,9 @@
#define ARM64_HAS_VIRT_HOST_EXTN 11
#define ARM64_WORKAROUND_CAVIUM_27456 12
#define ARM64_HAS_32BIT_EL0 13
+#define ARM64_HYP_OFFSET_LOW 14
-#define ARM64_NCAPS 14
+#define ARM64_NCAPS 15
#ifndef __ASSEMBLY__
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 12/55] arm64: KVM: Define HYP offset masks
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (10 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 11/55] arm64: Add ARM64_HYP_OFFSET_LOW capability Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 13/55] arm64: KVM: Refactor kern_hyp_va to deal with multiple offsets Marc Zyngier
` (43 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Define the two possible HYP VA regions in terms of VA_BITS,
and keep HYP_PAGE_OFFSET_MASK as a temporary compatibility
definition.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/kvm_mmu.h | 8 ++++++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 2f1e1ae..5e54323 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -68,8 +68,12 @@
* functionality is already mapped as part of the main kernel
* mappings, and none of this applies in that case.
*/
-#define HYP_PAGE_OFFSET_SHIFT VA_BITS
-#define HYP_PAGE_OFFSET_MASK ((UL(1) << HYP_PAGE_OFFSET_SHIFT) - 1)
+
+#define HYP_PAGE_OFFSET_HIGH_MASK ((UL(1) << VA_BITS) - 1)
+#define HYP_PAGE_OFFSET_LOW_MASK ((UL(1) << (VA_BITS - 1)) - 1)
+
+/* Temporary compat define */
+#define HYP_PAGE_OFFSET_MASK HYP_PAGE_OFFSET_HIGH_MASK
/*
* Our virtual mapping for the idmap-ed MMU-enable code. Must be
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 13/55] arm64: KVM: Refactor kern_hyp_va to deal with multiple offsets
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (11 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 12/55] arm64: KVM: Define HYP offset masks Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 14/55] arm/arm64: KVM: Export __hyp_text_start/end symbols Marc Zyngier
` (42 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
As we move towards a selectable HYP VA range, it is obvious that
we don't want to test a variable to find out if we need to use
the bottom VA range, the top VA range, or use the address as is
(for VHE).
Instead, we can expand our current helper to generate the right
mask or nop with code patching. We default to using the top VA
space, with alternatives to switch to the bottom one or to nop
out the instructions.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/kvm_hyp.h | 11 -----------
arch/arm64/include/asm/kvm_mmu.h | 42 +++++++++++++++++++++++++++++++++++++---
2 files changed, 39 insertions(+), 14 deletions(-)
diff --git a/arch/arm64/include/asm/kvm_hyp.h b/arch/arm64/include/asm/kvm_hyp.h
index 1d81f9a..cff5105 100644
--- a/arch/arm64/include/asm/kvm_hyp.h
+++ b/arch/arm64/include/asm/kvm_hyp.h
@@ -25,17 +25,6 @@
#define __hyp_text __section(.hyp.text) notrace
-static inline unsigned long __kern_hyp_va(unsigned long v)
-{
- asm volatile(ALTERNATIVE("and %0, %0, %1",
- "nop",
- ARM64_HAS_VIRT_HOST_EXTN)
- : "+r" (v) : "i" (HYP_PAGE_OFFSET_MASK));
- return v;
-}
-
-#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v)))
-
#define read_sysreg_elx(r,nvh,vh) \
({ \
u64 reg; \
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 5e54323..2970537 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -90,13 +90,33 @@
/*
* Convert a kernel VA into a HYP VA.
* reg: VA to be converted.
+ *
+ * This generates the following sequences:
+ * - High mask:
+ * and x0, x0, #HYP_PAGE_OFFSET_HIGH_MASK
+ * nop
+ * - Low mask:
+ * and x0, x0, #HYP_PAGE_OFFSET_HIGH_MASK
+ * and x0, x0, #HYP_PAGE_OFFSET_LOW_MASK
+ * - VHE:
+ * nop
+ * nop
+ *
+ * The "low mask" version works because the mask is a strict subset of
+ * the "high mask", hence performing the first mask for nothing.
+ * Should be completely invisible on any viable CPU.
*/
.macro kern_hyp_va reg
-alternative_if_not ARM64_HAS_VIRT_HOST_EXTN
- and \reg, \reg, #HYP_PAGE_OFFSET_MASK
+alternative_if_not ARM64_HAS_VIRT_HOST_EXTN
+ and \reg, \reg, #HYP_PAGE_OFFSET_HIGH_MASK
alternative_else
nop
alternative_endif
+alternative_if_not ARM64_HYP_OFFSET_LOW
+ nop
+alternative_else
+ and \reg, \reg, #HYP_PAGE_OFFSET_LOW_MASK
+alternative_endif
.endm
#else
@@ -107,7 +127,23 @@ alternative_endif
#include <asm/mmu_context.h>
#include <asm/pgtable.h>
-#define KERN_TO_HYP(kva) ((unsigned long)kva & HYP_PAGE_OFFSET_MASK)
+static inline unsigned long __kern_hyp_va(unsigned long v)
+{
+ asm volatile(ALTERNATIVE("and %0, %0, %1",
+ "nop",
+ ARM64_HAS_VIRT_HOST_EXTN)
+ : "+r" (v)
+ : "i" (HYP_PAGE_OFFSET_HIGH_MASK));
+ asm volatile(ALTERNATIVE("nop",
+ "and %0, %0, %1",
+ ARM64_HYP_OFFSET_LOW)
+ : "+r" (v)
+ : "i" (HYP_PAGE_OFFSET_LOW_MASK));
+ return v;
+}
+
+#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v)))
+#define KERN_TO_HYP(v) kern_hyp_va(v)
/*
* We currently only support a 40bit IPA.
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 14/55] arm/arm64: KVM: Export __hyp_text_start/end symbols
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (12 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 13/55] arm64: KVM: Refactor kern_hyp_va to deal with multiple offsets Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 15/55] arm64: KVM: Runtime detection of lower HYP offset Marc Zyngier
` (41 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Declare the __hyp_text_start/end symbols in asm/virt.h so that
they can be reused without having to declare them locally.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/virt.h | 4 ++++
arch/arm/kvm/mmu.c | 2 --
arch/arm64/include/asm/virt.h | 4 ++++
3 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/arch/arm/include/asm/virt.h b/arch/arm/include/asm/virt.h
index d4ceaf5..a2e75b8 100644
--- a/arch/arm/include/asm/virt.h
+++ b/arch/arm/include/asm/virt.h
@@ -80,6 +80,10 @@ static inline bool is_kernel_in_hyp_mode(void)
return false;
}
+/* The section containing the hypervisor idmap text */
+extern char __hyp_idmap_text_start[];
+extern char __hyp_idmap_text_end[];
+
/* The section containing the hypervisor text */
extern char __hyp_text_start[];
extern char __hyp_text_end[];
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 679608f..f004e70 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -32,8 +32,6 @@
#include "trace.h"
-extern char __hyp_idmap_text_start[], __hyp_idmap_text_end[];
-
static pgd_t *boot_hyp_pgd;
static pgd_t *hyp_pgd;
static pgd_t *merged_hyp_pgd;
diff --git a/arch/arm64/include/asm/virt.h b/arch/arm64/include/asm/virt.h
index dcbcf8d..88aa8ec 100644
--- a/arch/arm64/include/asm/virt.h
+++ b/arch/arm64/include/asm/virt.h
@@ -82,6 +82,10 @@ extern void verify_cpu_run_el(void);
static inline void verify_cpu_run_el(void) {}
#endif
+/* The section containing the hypervisor idmap text */
+extern char __hyp_idmap_text_start[];
+extern char __hyp_idmap_text_end[];
+
/* The section containing the hypervisor text */
extern char __hyp_text_start[];
extern char __hyp_text_end[];
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 15/55] arm64: KVM: Runtime detection of lower HYP offset
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (13 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 14/55] arm/arm64: KVM: Export __hyp_text_start/end symbols Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 16/55] arm/arm64: KVM: Always have merged page tables Marc Zyngier
` (40 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Add the code that enables the switch to the lower HYP VA range.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/kernel/cpufeature.c | 19 +++++++++++++++++++
1 file changed, 19 insertions(+)
diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c
index 811773d..ffb3e14 100644
--- a/arch/arm64/kernel/cpufeature.c
+++ b/arch/arm64/kernel/cpufeature.c
@@ -726,6 +726,19 @@ static bool runs_at_el2(const struct arm64_cpu_capabilities *entry, int __unused
return is_kernel_in_hyp_mode();
}
+static bool hyp_offset_low(const struct arm64_cpu_capabilities *entry,
+ int __unused)
+{
+ phys_addr_t idmap_addr = virt_to_phys(__hyp_idmap_text_start);
+
+ /*
+ * Activate the lower HYP offset only if:
+ * - the idmap doesn't clash with it,
+ * - the kernel is not running at EL2.
+ */
+ return idmap_addr > GENMASK(VA_BITS - 2, 0) && !is_kernel_in_hyp_mode();
+}
+
static const struct arm64_cpu_capabilities arm64_features[] = {
{
.desc = "GIC system register CPU interface",
@@ -803,6 +816,12 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
.field_pos = ID_AA64PFR0_EL0_SHIFT,
.min_field_value = ID_AA64PFR0_EL0_32BIT_64BIT,
},
+ {
+ .desc = "Reduced HYP mapping offset",
+ .capability = ARM64_HYP_OFFSET_LOW,
+ .def_scope = SCOPE_SYSTEM,
+ .matches = hyp_offset_low,
+ },
{},
};
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 16/55] arm/arm64: KVM: Always have merged page tables
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (14 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 15/55] arm64: KVM: Runtime detection of lower HYP offset Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 17/55] arm64: KVM: Simplify HYP init/teardown Marc Zyngier
` (39 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
We're in a position where we can now always have "merged" page
tables, where both the runtime mapping and the idmap coexist.
This results in some code being removed, but there is more to come.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/kvm/mmu.c | 74 +++++++++++++++++++++++---------------------------
arch/arm64/kvm/reset.c | 31 +++++----------------
2 files changed, 41 insertions(+), 64 deletions(-)
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index f004e70..80d3737 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -492,13 +492,12 @@ void free_boot_hyp_pgd(void)
if (boot_hyp_pgd) {
unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE);
- unmap_hyp_range(boot_hyp_pgd, TRAMPOLINE_VA, PAGE_SIZE);
free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order);
boot_hyp_pgd = NULL;
}
if (hyp_pgd)
- unmap_hyp_range(hyp_pgd, TRAMPOLINE_VA, PAGE_SIZE);
+ unmap_hyp_range(hyp_pgd, hyp_idmap_start, PAGE_SIZE);
mutex_unlock(&kvm_hyp_pgd_mutex);
}
@@ -1691,7 +1690,7 @@ phys_addr_t kvm_mmu_get_boot_httbr(void)
if (__kvm_cpu_uses_extended_idmap())
return virt_to_phys(merged_hyp_pgd);
else
- return virt_to_phys(boot_hyp_pgd);
+ return virt_to_phys(hyp_pgd);
}
phys_addr_t kvm_get_idmap_vector(void)
@@ -1704,6 +1703,22 @@ phys_addr_t kvm_get_idmap_start(void)
return hyp_idmap_start;
}
+static int kvm_map_idmap_text(pgd_t *pgd)
+{
+ int err;
+
+ /* Create the idmap in the boot page tables */
+ err = __create_hyp_mappings(pgd,
+ hyp_idmap_start, hyp_idmap_end,
+ __phys_to_pfn(hyp_idmap_start),
+ PAGE_HYP_EXEC);
+ if (err)
+ kvm_err("Failed to idmap %lx-%lx\n",
+ hyp_idmap_start, hyp_idmap_end);
+
+ return err;
+}
+
int kvm_mmu_init(void)
{
int err;
@@ -1719,27 +1734,25 @@ int kvm_mmu_init(void)
BUG_ON((hyp_idmap_start ^ (hyp_idmap_end - 1)) & PAGE_MASK);
hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order);
- boot_hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order);
-
- if (!hyp_pgd || !boot_hyp_pgd) {
+ if (!hyp_pgd) {
kvm_err("Hyp mode PGD not allocated\n");
err = -ENOMEM;
goto out;
}
- /* Create the idmap in the boot page tables */
- err = __create_hyp_mappings(boot_hyp_pgd,
- hyp_idmap_start, hyp_idmap_end,
- __phys_to_pfn(hyp_idmap_start),
- PAGE_HYP_EXEC);
+ if (__kvm_cpu_uses_extended_idmap()) {
+ boot_hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
+ hyp_pgd_order);
+ if (!boot_hyp_pgd) {
+ kvm_err("Hyp boot PGD not allocated\n");
+ err = -ENOMEM;
+ goto out;
+ }
- if (err) {
- kvm_err("Failed to idmap %lx-%lx\n",
- hyp_idmap_start, hyp_idmap_end);
- goto out;
- }
+ err = kvm_map_idmap_text(boot_hyp_pgd);
+ if (err)
+ goto out;
- if (__kvm_cpu_uses_extended_idmap()) {
merged_hyp_pgd = (pgd_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
if (!merged_hyp_pgd) {
kvm_err("Failed to allocate extra HYP pgd\n");
@@ -1747,29 +1760,10 @@ int kvm_mmu_init(void)
}
__kvm_extend_hypmap(boot_hyp_pgd, hyp_pgd, merged_hyp_pgd,
hyp_idmap_start);
- return 0;
- }
-
- /* Map the very same page at the trampoline VA */
- err = __create_hyp_mappings(boot_hyp_pgd,
- TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE,
- __phys_to_pfn(hyp_idmap_start),
- PAGE_HYP_EXEC);
- if (err) {
- kvm_err("Failed to map trampoline @%lx into boot HYP pgd\n",
- TRAMPOLINE_VA);
- goto out;
- }
-
- /* Map the same page again into the runtime page tables */
- err = __create_hyp_mappings(hyp_pgd,
- TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE,
- __phys_to_pfn(hyp_idmap_start),
- PAGE_HYP_EXEC);
- if (err) {
- kvm_err("Failed to map trampoline @%lx into runtime HYP pgd\n",
- TRAMPOLINE_VA);
- goto out;
+ } else {
+ err = kvm_map_idmap_text(hyp_pgd);
+ if (err)
+ goto out;
}
return 0;
diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
index 7be24f2..8ed7e4a 100644
--- a/arch/arm64/kvm/reset.c
+++ b/arch/arm64/kvm/reset.c
@@ -133,30 +133,13 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu)
return kvm_timer_vcpu_reset(vcpu, cpu_vtimer_irq);
}
-extern char __hyp_idmap_text_start[];
-
unsigned long kvm_hyp_reset_entry(void)
{
- if (!__kvm_cpu_uses_extended_idmap()) {
- unsigned long offset;
-
- /*
- * Find the address of __kvm_hyp_reset() in the trampoline page.
- * This is present in the running page tables, and the boot page
- * tables, so we call the code here to start the trampoline
- * dance in reverse.
- */
- offset = (unsigned long)__kvm_hyp_reset
- - ((unsigned long)__hyp_idmap_text_start & PAGE_MASK);
-
- return TRAMPOLINE_VA + offset;
- } else {
- /*
- * KVM is running with merged page tables, which don't have the
- * trampoline page mapped. We know the idmap is still mapped,
- * but can't be called into directly. Use
- * __extended_idmap_trampoline to do the call.
- */
- return (unsigned long)kvm_ksym_ref(__extended_idmap_trampoline);
- }
+ /*
+ * KVM is running with merged page tables, which don't have the
+ * trampoline page mapped. We know the idmap is still mapped,
+ * but can't be called into directly. Use
+ * __extended_idmap_trampoline to do the call.
+ */
+ return (unsigned long)kvm_ksym_ref(__extended_idmap_trampoline);
}
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 17/55] arm64: KVM: Simplify HYP init/teardown
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (15 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 16/55] arm/arm64: KVM: Always have merged page tables Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 18/55] arm/arm64: KVM: Drop boot_pgd Marc Zyngier
` (38 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Now that we only have the "merged page tables" case to deal with,
there is a bunch of things we can simplify in the HYP code (both
at init and teardown time).
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm64/include/asm/kvm_host.h | 12 ++------
arch/arm64/kvm/hyp-init.S | 61 +++++----------------------------------
arch/arm64/kvm/hyp/entry.S | 19 ------------
arch/arm64/kvm/hyp/hyp-entry.S | 15 ++++++++++
arch/arm64/kvm/reset.c | 11 -------
5 files changed, 26 insertions(+), 92 deletions(-)
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 49095fc..88462c3 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -48,7 +48,6 @@
int __attribute_const__ kvm_target_cpu(void);
int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
int kvm_arch_dev_ioctl_check_extension(long ext);
-unsigned long kvm_hyp_reset_entry(void);
void __extended_idmap_trampoline(phys_addr_t boot_pgd, phys_addr_t idmap_start);
struct kvm_arch {
@@ -357,19 +356,14 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
* Call initialization code, and switch to the full blown
* HYP code.
*/
- __kvm_call_hyp((void *)boot_pgd_ptr, pgd_ptr,
- hyp_stack_ptr, vector_ptr);
+ __kvm_call_hyp((void *)pgd_ptr, hyp_stack_ptr, vector_ptr);
}
+void __kvm_hyp_teardown(void);
static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr,
phys_addr_t phys_idmap_start)
{
- /*
- * Call reset code, and switch back to stub hyp vectors.
- * Uses __kvm_call_hyp() to avoid kaslr's kvm_ksym_ref() translation.
- */
- __kvm_call_hyp((void *)kvm_hyp_reset_entry(),
- boot_pgd_ptr, phys_idmap_start);
+ kvm_call_hyp(__kvm_hyp_teardown, phys_idmap_start);
}
static inline void kvm_arch_hardware_unsetup(void) {}
diff --git a/arch/arm64/kvm/hyp-init.S b/arch/arm64/kvm/hyp-init.S
index a873a6d..6b29d3d 100644
--- a/arch/arm64/kvm/hyp-init.S
+++ b/arch/arm64/kvm/hyp-init.S
@@ -53,10 +53,9 @@ __invalid:
b .
/*
- * x0: HYP boot pgd
- * x1: HYP pgd
- * x2: HYP stack
- * x3: HYP vectors
+ * x0: HYP pgd
+ * x1: HYP stack
+ * x2: HYP vectors
*/
__do_hyp_init:
@@ -110,71 +109,27 @@ __do_hyp_init:
msr sctlr_el2, x4
isb
- /* Skip the trampoline dance if we merged the boot and runtime PGDs */
- cmp x0, x1
- b.eq merged
-
- /* MMU is now enabled. Get ready for the trampoline dance */
- ldr x4, =TRAMPOLINE_VA
- adr x5, target
- bfi x4, x5, #0, #PAGE_SHIFT
- br x4
-
-target: /* We're now in the trampoline code, switch page tables */
- msr ttbr0_el2, x1
- isb
-
- /* Invalidate the old TLBs */
- tlbi alle2
- dsb sy
-
-merged:
/* Set the stack and new vectors */
+ kern_hyp_va x1
+ mov sp, x1
kern_hyp_va x2
- mov sp, x2
- kern_hyp_va x3
- msr vbar_el2, x3
+ msr vbar_el2, x2
/* Hello, World! */
eret
ENDPROC(__kvm_hyp_init)
/*
- * Reset kvm back to the hyp stub. This is the trampoline dance in
- * reverse. If kvm used an extended idmap, __extended_idmap_trampoline
- * calls this code directly in the idmap. In this case switching to the
- * boot tables is a no-op.
- *
- * x0: HYP boot pgd
- * x1: HYP phys_idmap_start
+ * Reset kvm back to the hyp stub.
*/
ENTRY(__kvm_hyp_reset)
- /* We're in trampoline code in VA, switch back to boot page tables */
- msr ttbr0_el2, x0
- isb
-
- /* Ensure the PA branch doesn't find a stale tlb entry or stale code. */
- ic iallu
- tlbi alle2
- dsb sy
- isb
-
- /* Branch into PA space */
- adr x0, 1f
- bfi x1, x0, #0, #PAGE_SHIFT
- br x1
-
/* We're now in idmap, disable MMU */
-1: mrs x0, sctlr_el2
+ mrs x0, sctlr_el2
ldr x1, =SCTLR_ELx_FLAGS
bic x0, x0, x1 // Clear SCTL_M and etc
msr sctlr_el2, x0
isb
- /* Invalidate the old TLBs */
- tlbi alle2
- dsb sy
-
/* Install stub vectors */
adr_l x0, __hyp_stub_vectors
msr vbar_el2, x0
diff --git a/arch/arm64/kvm/hyp/entry.S b/arch/arm64/kvm/hyp/entry.S
index 70254a6..ce9e5e5 100644
--- a/arch/arm64/kvm/hyp/entry.S
+++ b/arch/arm64/kvm/hyp/entry.S
@@ -164,22 +164,3 @@ alternative_endif
eret
ENDPROC(__fpsimd_guest_restore)
-
-/*
- * When using the extended idmap, we don't have a trampoline page we can use
- * while we switch pages tables during __kvm_hyp_reset. Accessing the idmap
- * directly would be ideal, but if we're using the extended idmap then the
- * idmap is located above HYP_PAGE_OFFSET, and the address will be masked by
- * kvm_call_hyp using kern_hyp_va.
- *
- * x0: HYP boot pgd
- * x1: HYP phys_idmap_start
- */
-ENTRY(__extended_idmap_trampoline)
- mov x4, x1
- adr_l x3, __kvm_hyp_reset
-
- /* insert __kvm_hyp_reset()s offset into phys_idmap_start */
- bfi x4, x3, #0, #PAGE_SHIFT
- br x4
-ENDPROC(__extended_idmap_trampoline)
diff --git a/arch/arm64/kvm/hyp/hyp-entry.S b/arch/arm64/kvm/hyp/hyp-entry.S
index 2d87f36..f6d9694 100644
--- a/arch/arm64/kvm/hyp/hyp-entry.S
+++ b/arch/arm64/kvm/hyp/hyp-entry.S
@@ -62,6 +62,21 @@ ENTRY(__vhe_hyp_call)
isb
ret
ENDPROC(__vhe_hyp_call)
+
+/*
+ * Compute the idmap address of __kvm_hyp_reset based on the idmap
+ * start passed as a parameter, and jump there.
+ *
+ * x0: HYP phys_idmap_start
+ */
+ENTRY(__kvm_hyp_teardown)
+ mov x4, x0
+ adr_l x3, __kvm_hyp_reset
+
+ /* insert __kvm_hyp_reset()s offset into phys_idmap_start */
+ bfi x4, x3, #0, #PAGE_SHIFT
+ br x4
+ENDPROC(__kvm_hyp_teardown)
el1_sync: // Guest trapped into EL2
save_x0_to_x3
diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
index 8ed7e4a..79f3248 100644
--- a/arch/arm64/kvm/reset.c
+++ b/arch/arm64/kvm/reset.c
@@ -132,14 +132,3 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu)
/* Reset timer */
return kvm_timer_vcpu_reset(vcpu, cpu_vtimer_irq);
}
-
-unsigned long kvm_hyp_reset_entry(void)
-{
- /*
- * KVM is running with merged page tables, which don't have the
- * trampoline page mapped. We know the idmap is still mapped,
- * but can't be called into directly. Use
- * __extended_idmap_trampoline to do the call.
- */
- return (unsigned long)kvm_ksym_ref(__extended_idmap_trampoline);
-}
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 18/55] arm/arm64: KVM: Drop boot_pgd
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (16 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 17/55] arm64: KVM: Simplify HYP init/teardown Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 19/55] arm/arm64: KVM: Kill free_boot_hyp_pgd Marc Zyngier
` (37 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Since we now only have one set of page tables, the concept of
boot_pgd is useless and can be removed. We still keep it as
an element of the "extended idmap" thing.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_host.h | 8 +++-----
arch/arm/include/asm/kvm_mmu.h | 1 -
arch/arm/kvm/arm.c | 15 +++------------
arch/arm/kvm/mmu.c | 8 --------
arch/arm64/include/asm/kvm_host.h | 6 ++----
arch/arm64/include/asm/kvm_mmu.h | 1 -
6 files changed, 8 insertions(+), 31 deletions(-)
diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
index 96387d4..020f4eb 100644
--- a/arch/arm/include/asm/kvm_host.h
+++ b/arch/arm/include/asm/kvm_host.h
@@ -241,8 +241,7 @@ int kvm_arm_coproc_set_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *);
int handle_exit(struct kvm_vcpu *vcpu, struct kvm_run *run,
int exception_index);
-static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
- phys_addr_t pgd_ptr,
+static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
unsigned long hyp_stack_ptr,
unsigned long vector_ptr)
{
@@ -272,12 +271,11 @@ static inline void __cpu_init_stage2(void)
kvm_call_hyp(__init_stage2_translation);
}
-static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr,
- phys_addr_t phys_idmap_start)
+static inline void __cpu_reset_hyp_mode(phys_addr_t phys_idmap_start)
{
/*
* TODO
- * kvm_call_reset(boot_pgd_ptr, phys_idmap_start);
+ * kvm_call_reset(phys_idmap_start);
*/
}
diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h
index 6cb4d4d..5d161d1 100644
--- a/arch/arm/include/asm/kvm_mmu.h
+++ b/arch/arm/include/asm/kvm_mmu.h
@@ -65,7 +65,6 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run);
void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu);
phys_addr_t kvm_mmu_get_httbr(void);
-phys_addr_t kvm_mmu_get_boot_httbr(void);
phys_addr_t kvm_get_idmap_vector(void);
phys_addr_t kvm_get_idmap_start(void);
int kvm_mmu_init(void);
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index c74483f..0887cc1 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -1038,7 +1038,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
static void cpu_init_hyp_mode(void *dummy)
{
- phys_addr_t boot_pgd_ptr;
phys_addr_t pgd_ptr;
unsigned long hyp_stack_ptr;
unsigned long stack_page;
@@ -1047,13 +1046,12 @@ static void cpu_init_hyp_mode(void *dummy)
/* Switch from the HYP stub to our own HYP init vector */
__hyp_set_vectors(kvm_get_idmap_vector());
- boot_pgd_ptr = kvm_mmu_get_boot_httbr();
pgd_ptr = kvm_mmu_get_httbr();
stack_page = __this_cpu_read(kvm_arm_hyp_stack_page);
hyp_stack_ptr = stack_page + PAGE_SIZE;
vector_ptr = (unsigned long)kvm_ksym_ref(__kvm_hyp_vector);
- __cpu_init_hyp_mode(boot_pgd_ptr, pgd_ptr, hyp_stack_ptr, vector_ptr);
+ __cpu_init_hyp_mode(pgd_ptr, hyp_stack_ptr, vector_ptr);
__cpu_init_stage2();
kvm_arm_init_debug();
@@ -1075,15 +1073,8 @@ static void cpu_hyp_reinit(void)
static void cpu_hyp_reset(void)
{
- phys_addr_t boot_pgd_ptr;
- phys_addr_t phys_idmap_start;
-
- if (!is_kernel_in_hyp_mode()) {
- boot_pgd_ptr = kvm_mmu_get_boot_httbr();
- phys_idmap_start = kvm_get_idmap_start();
-
- __cpu_reset_hyp_mode(boot_pgd_ptr, phys_idmap_start);
- }
+ if (!is_kernel_in_hyp_mode())
+ __cpu_reset_hyp_mode(kvm_get_idmap_start());
}
static void _kvm_arch_hardware_enable(void *discard)
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 80d3737..dd4ccc7 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -1685,14 +1685,6 @@ phys_addr_t kvm_mmu_get_httbr(void)
return virt_to_phys(hyp_pgd);
}
-phys_addr_t kvm_mmu_get_boot_httbr(void)
-{
- if (__kvm_cpu_uses_extended_idmap())
- return virt_to_phys(merged_hyp_pgd);
- else
- return virt_to_phys(hyp_pgd);
-}
-
phys_addr_t kvm_get_idmap_vector(void)
{
return hyp_idmap_vector;
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 88462c3..6731d4e 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -347,8 +347,7 @@ int kvm_perf_teardown(void);
struct kvm_vcpu *kvm_mpidr_to_vcpu(struct kvm *kvm, unsigned long mpidr);
-static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
- phys_addr_t pgd_ptr,
+static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
unsigned long hyp_stack_ptr,
unsigned long vector_ptr)
{
@@ -360,8 +359,7 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr,
}
void __kvm_hyp_teardown(void);
-static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr,
- phys_addr_t phys_idmap_start)
+static inline void __cpu_reset_hyp_mode(phys_addr_t phys_idmap_start)
{
kvm_call_hyp(__kvm_hyp_teardown, phys_idmap_start);
}
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 2970537..390acab 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -170,7 +170,6 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run);
void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu);
phys_addr_t kvm_mmu_get_httbr(void);
-phys_addr_t kvm_mmu_get_boot_httbr(void);
phys_addr_t kvm_get_idmap_vector(void);
phys_addr_t kvm_get_idmap_start(void);
int kvm_mmu_init(void);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 19/55] arm/arm64: KVM: Kill free_boot_hyp_pgd
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (17 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 18/55] arm/arm64: KVM: Drop boot_pgd Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 20/55] arm: KVM: Simplify HYP init Marc Zyngier
` (36 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
There is no way to free the boot PGD, because it doesn't exist
anymore as a standalone entity.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_mmu.h | 1 -
arch/arm/kvm/arm.c | 4 ----
arch/arm/kvm/mmu.c | 30 +++++++-----------------------
arch/arm64/include/asm/kvm_mmu.h | 1 -
4 files changed, 7 insertions(+), 29 deletions(-)
diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h
index 5d161d1..d5fd9fb 100644
--- a/arch/arm/include/asm/kvm_mmu.h
+++ b/arch/arm/include/asm/kvm_mmu.h
@@ -51,7 +51,6 @@
int create_hyp_mappings(void *from, void *to, pgprot_t prot);
int create_hyp_io_mappings(void *from, void *to, phys_addr_t);
-void free_boot_hyp_pgd(void);
void free_hyp_pgds(void);
void stage2_unmap_vm(struct kvm *kvm);
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 0887cc1..9b8c537 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -1323,10 +1323,6 @@ static int init_hyp_mode(void)
}
}
-#ifndef CONFIG_HOTPLUG_CPU
- free_boot_hyp_pgd();
-#endif
-
/* set size of VMID supported by CPU */
kvm_vmid_bits = kvm_get_vmid_bits();
kvm_info("%d-bit VMID\n", kvm_vmid_bits);
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index dd4ccc7..0b36dd52 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -482,27 +482,6 @@ static void unmap_hyp_range(pgd_t *pgdp, phys_addr_t start, u64 size)
}
/**
- * free_boot_hyp_pgd - free HYP boot page tables
- *
- * Free the HYP boot page tables. The bounce page is also freed.
- */
-void free_boot_hyp_pgd(void)
-{
- mutex_lock(&kvm_hyp_pgd_mutex);
-
- if (boot_hyp_pgd) {
- unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE);
- free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order);
- boot_hyp_pgd = NULL;
- }
-
- if (hyp_pgd)
- unmap_hyp_range(hyp_pgd, hyp_idmap_start, PAGE_SIZE);
-
- mutex_unlock(&kvm_hyp_pgd_mutex);
-}
-
-/**
* free_hyp_pgds - free Hyp-mode page tables
*
* Assumes hyp_pgd is a page table used strictly in Hyp-mode and
@@ -516,11 +495,16 @@ void free_hyp_pgds(void)
{
unsigned long addr;
- free_boot_hyp_pgd();
-
mutex_lock(&kvm_hyp_pgd_mutex);
+ if (boot_hyp_pgd) {
+ unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE);
+ free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order);
+ boot_hyp_pgd = NULL;
+ }
+
if (hyp_pgd) {
+ unmap_hyp_range(hyp_pgd, hyp_idmap_start, PAGE_SIZE);
for (addr = PAGE_OFFSET; virt_addr_valid(addr); addr += PGDIR_SIZE)
unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE);
for (addr = VMALLOC_START; is_vmalloc_addr((void*)addr); addr += PGDIR_SIZE)
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 390acab..b89122e 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -156,7 +156,6 @@ static inline unsigned long __kern_hyp_va(unsigned long v)
int create_hyp_mappings(void *from, void *to, pgprot_t prot);
int create_hyp_io_mappings(void *from, void *to, phys_addr_t);
-void free_boot_hyp_pgd(void);
void free_hyp_pgds(void);
void stage2_unmap_vm(struct kvm *kvm);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 20/55] arm: KVM: Simplify HYP init
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (18 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 19/55] arm/arm64: KVM: Kill free_boot_hyp_pgd Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 21/55] arm: KVM: Allow hyp teardown Marc Zyngier
` (35 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
Just like for arm64, we can now make the HYP setup a lot simpler,
and we can now initialise it in one go (instead of the two
phases we currently have).
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_host.h | 15 +++++--------
arch/arm/kvm/init.S | 49 ++++++++---------------------------------
2 files changed, 14 insertions(+), 50 deletions(-)
diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
index 020f4eb..eafbfd5 100644
--- a/arch/arm/include/asm/kvm_host.h
+++ b/arch/arm/include/asm/kvm_host.h
@@ -250,18 +250,13 @@ static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
* code. The init code doesn't need to preserve these
* registers as r0-r3 are already callee saved according to
* the AAPCS.
- * Note that we slightly misuse the prototype by casing the
+ * Note that we slightly misuse the prototype by casting the
* stack pointer to a void *.
- *
- * We don't have enough registers to perform the full init in
- * one go. Install the boot PGD first, and then install the
- * runtime PGD, stack pointer and vectors. The PGDs are always
- * passed as the third argument, in order to be passed into
- * r2-r3 to the init code (yes, this is compliant with the
- * PCS!).
- */
- kvm_call_hyp(NULL, 0, boot_pgd_ptr);
+ * The PGDs are always passed as the third argument, in order
+ * to be passed into r2-r3 to the init code (yes, this is
+ * compliant with the PCS!).
+ */
kvm_call_hyp((void*)hyp_stack_ptr, vector_ptr, pgd_ptr);
}
diff --git a/arch/arm/kvm/init.S b/arch/arm/kvm/init.S
index 1f9ae17..b82a99d 100644
--- a/arch/arm/kvm/init.S
+++ b/arch/arm/kvm/init.S
@@ -32,23 +32,13 @@
* r2,r3 = Hypervisor pgd pointer
*
* The init scenario is:
- * - We jump in HYP with four parameters: boot HYP pgd, runtime HYP pgd,
- * runtime stack, runtime vectors
- * - Enable the MMU with the boot pgd
- * - Jump to a target into the trampoline page (remember, this is the same
- * physical page!)
- * - Now switch to the runtime pgd (same VA, and still the same physical
- * page!)
+ * - We jump in HYP with 3 parameters: runtime HYP pgd, runtime stack,
+ * runtime vectors
* - Invalidate TLBs
* - Set stack and vectors
+ * - Setup the page tables
+ * - Enable the MMU
* - Profit! (or eret, if you only care about the code).
- *
- * As we only have four registers available to pass parameters (and we
- * need six), we split the init in two phases:
- * - Phase 1: r0 = 0, r1 = 0, r2,r3 contain the boot PGD.
- * Provides the basic HYP init, and enable the MMU.
- * - Phase 2: r0 = ToS, r1 = vectors, r2,r3 contain the runtime PGD.
- * Switches to the runtime PGD, set stack and vectors.
*/
.text
@@ -68,8 +58,11 @@ __kvm_hyp_init:
W(b) .
__do_hyp_init:
- cmp r0, #0 @ We have a SP?
- bne phase2 @ Yes, second stage init
+ @ Set stack pointer
+ mov sp, r0
+
+ @ Set HVBAR to point to the HYP vectors
+ mcr p15, 4, r1, c12, c0, 0 @ HVBAR
@ Set the HTTBR to point to the hypervisor PGD pointer passed
mcrr p15, 4, rr_lo_hi(r2, r3), c2
@@ -114,33 +107,9 @@ __do_hyp_init:
THUMB( ldr r2, =(HSCTLR_M | HSCTLR_A | HSCTLR_TE) )
orr r1, r1, r2
orr r0, r0, r1
- isb
mcr p15, 4, r0, c1, c0, 0 @ HSCR
-
- @ End of init phase-1
- eret
-
-phase2:
- @ Set stack pointer
- mov sp, r0
-
- @ Set HVBAR to point to the HYP vectors
- mcr p15, 4, r1, c12, c0, 0 @ HVBAR
-
- @ Jump to the trampoline page
- ldr r0, =TRAMPOLINE_VA
- adr r1, target
- bfi r0, r1, #0, #PAGE_SHIFT
- ret r0
-
-target: @ We're now in the trampoline code, switch page tables
- mcrr p15, 4, rr_lo_hi(r2, r3), c2
isb
- @ Invalidate the old TLBs
- mcr p15, 4, r0, c8, c7, 0 @ TLBIALLH
- dsb ish
-
eret
.ltorg
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 21/55] arm: KVM: Allow hyp teardown
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (19 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 20/55] arm: KVM: Simplify HYP init Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 22/55] arm/arm64: KVM: Prune unused #defines Marc Zyngier
` (34 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
So far, KVM was getting in the way of kexec on 32bit (and the arm64
kexec hackers couldn't be bothered to fix it on 32bit...).
With simpler page tables, tearing KVM down becomes very easy, so
let's just do it.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_asm.h | 2 ++
arch/arm/include/asm/kvm_host.h | 8 +++-----
arch/arm/kvm/arm.c | 3 ++-
arch/arm/kvm/init.S | 15 +++++++++++++++
arch/arm64/include/asm/kvm_host.h | 3 ++-
5 files changed, 24 insertions(+), 7 deletions(-)
diff --git a/arch/arm/include/asm/kvm_asm.h b/arch/arm/include/asm/kvm_asm.h
index 3d5a5cd..58faff5 100644
--- a/arch/arm/include/asm/kvm_asm.h
+++ b/arch/arm/include/asm/kvm_asm.h
@@ -66,6 +66,8 @@ extern void __kvm_tlb_flush_vmid(struct kvm *kvm);
extern int __kvm_vcpu_run(struct kvm_vcpu *vcpu);
extern void __init_stage2_translation(void);
+
+extern void __kvm_hyp_reset(unsigned long);
#endif
#endif /* __ARM_KVM_ASM_H__ */
diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
index eafbfd5..58d0b69 100644
--- a/arch/arm/include/asm/kvm_host.h
+++ b/arch/arm/include/asm/kvm_host.h
@@ -266,12 +266,10 @@ static inline void __cpu_init_stage2(void)
kvm_call_hyp(__init_stage2_translation);
}
-static inline void __cpu_reset_hyp_mode(phys_addr_t phys_idmap_start)
+static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr,
+ phys_addr_t phys_idmap_start)
{
- /*
- * TODO
- * kvm_call_reset(phys_idmap_start);
- */
+ kvm_call_hyp((void *)virt_to_idmap(__kvm_hyp_reset), vector_ptr);
}
static inline int kvm_arch_dev_ioctl_check_extension(long ext)
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 9b8c537..7cf266c 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -1074,7 +1074,8 @@ static void cpu_hyp_reinit(void)
static void cpu_hyp_reset(void)
{
if (!is_kernel_in_hyp_mode())
- __cpu_reset_hyp_mode(kvm_get_idmap_start());
+ __cpu_reset_hyp_mode(hyp_default_vectors,
+ kvm_get_idmap_start());
}
static void _kvm_arch_hardware_enable(void *discard)
diff --git a/arch/arm/kvm/init.S b/arch/arm/kvm/init.S
index b82a99d..bf89c91 100644
--- a/arch/arm/kvm/init.S
+++ b/arch/arm/kvm/init.S
@@ -112,6 +112,21 @@ __do_hyp_init:
eret
+ @ r0 : stub vectors address
+ENTRY(__kvm_hyp_reset)
+ /* We're now in idmap, disable MMU */
+ mrc p15, 4, r1, c1, c0, 0 @ HSCTLR
+ ldr r2, =(HSCTLR_M | HSCTLR_A | HSCTLR_C | HSCTLR_I)
+ bic r1, r1, r2
+ mcr p15, 4, r1, c1, c0, 0 @ HSCTLR
+
+ /* Install stub vectors */
+ mcr p15, 4, r0, c12, c0, 0 @ HVBAR
+ isb
+
+ eret
+ENDPROC(__kvm_hyp_reset)
+
.ltorg
.globl __kvm_hyp_init_end
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 6731d4e..69d5cc2d 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -359,7 +359,8 @@ static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr,
}
void __kvm_hyp_teardown(void);
-static inline void __cpu_reset_hyp_mode(phys_addr_t phys_idmap_start)
+static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr,
+ phys_addr_t phys_idmap_start)
{
kvm_call_hyp(__kvm_hyp_teardown, phys_idmap_start);
}
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 22/55] arm/arm64: KVM: Prune unused #defines
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (20 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 21/55] arm: KVM: Allow hyp teardown Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 23/55] arm/arm64: KVM: Check that IDMAP doesn't intersect with VA range Marc Zyngier
` (33 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
We can now remove a number of dead #defines, thanks to the trampoline
code being gone.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_mmu.h | 9 ---------
arch/arm64/include/asm/kvm_mmu.h | 10 ----------
2 files changed, 19 deletions(-)
diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h
index d5fd9fb..73c2818 100644
--- a/arch/arm/include/asm/kvm_mmu.h
+++ b/arch/arm/include/asm/kvm_mmu.h
@@ -26,18 +26,9 @@
* We directly use the kernel VA for the HYP, as we can directly share
* the mapping (HTTBR "covers" TTBR1).
*/
-#define HYP_PAGE_OFFSET_MASK UL(~0)
-#define HYP_PAGE_OFFSET PAGE_OFFSET
#define KERN_TO_HYP(kva) (kva)
/*
- * Our virtual mapping for the boot-time MMU-enable code. Must be
- * shared across all the page-tables. Conveniently, we use the vectors
- * page, where no kernel data will ever be shared with HYP.
- */
-#define TRAMPOLINE_VA UL(CONFIG_VECTORS_BASE)
-
-/*
* KVM_MMU_CACHE_MIN_PAGES is the number of stage2 page table translation levels.
*/
#define KVM_MMU_CACHE_MIN_PAGES 2
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index b89122e..9226f8b 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -72,16 +72,6 @@
#define HYP_PAGE_OFFSET_HIGH_MASK ((UL(1) << VA_BITS) - 1)
#define HYP_PAGE_OFFSET_LOW_MASK ((UL(1) << (VA_BITS - 1)) - 1)
-/* Temporary compat define */
-#define HYP_PAGE_OFFSET_MASK HYP_PAGE_OFFSET_HIGH_MASK
-
-/*
- * Our virtual mapping for the idmap-ed MMU-enable code. Must be
- * shared across all the page-tables. Conveniently, we use the last
- * possible page, where no kernel mapping will ever exist.
- */
-#define TRAMPOLINE_VA (HYP_PAGE_OFFSET_MASK & PAGE_MASK)
-
#ifdef __ASSEMBLY__
#include <asm/alternative.h>
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 23/55] arm/arm64: KVM: Check that IDMAP doesn't intersect with VA range
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (21 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 22/55] arm/arm64: KVM: Prune unused #defines Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 24/55] arm/arm64: Get rid of KERN_TO_HYP Marc Zyngier
` (32 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
This is more of a safety measure than anything else: If we end-up
with an idmap page that intersect with the range picked for the
the HYP VA space, abort the KVM setup, as it is unsafe to go
further.
I cannot imagine it happening on 64bit (we have a mechanism to
work around it), but could potentially occur on a 32bit system with
the kernel loaded high enough in memory so that in conflicts with
the kernel VA.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/kvm/mmu.c | 15 +++++++++++++++
1 file changed, 15 insertions(+)
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 0b36dd52..8a0aa37 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -1709,6 +1709,21 @@ int kvm_mmu_init(void)
*/
BUG_ON((hyp_idmap_start ^ (hyp_idmap_end - 1)) & PAGE_MASK);
+ kvm_info("IDMAP page: %lx\n", hyp_idmap_start);
+ kvm_info("HYP VA range: %lx:%lx\n",
+ KERN_TO_HYP(PAGE_OFFSET), KERN_TO_HYP(~0UL));
+
+ if (hyp_idmap_start >= KERN_TO_HYP(PAGE_OFFSET) &&
+ hyp_idmap_start < KERN_TO_HYP(~0UL)) {
+ /*
+ * The idmap page is intersecting with the VA space,
+ * it is not safe to continue further.
+ */
+ kvm_err("IDMAP intersecting with HYP VA, unable to continue\n");
+ err = -EINVAL;
+ goto out;
+ }
+
hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order);
if (!hyp_pgd) {
kvm_err("Hyp mode PGD not allocated\n");
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 24/55] arm/arm64: Get rid of KERN_TO_HYP
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (22 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 23/55] arm/arm64: KVM: Check that IDMAP doesn't intersect with VA range Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 25/55] arm64: KVM: Clean up a condition Marc Zyngier
` (31 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
We have both KERN_TO_HYP and kern_hyp_va, which do the exact same
thing. Let's standardize on the latter.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
arch/arm/include/asm/kvm_hyp.h | 2 --
arch/arm/include/asm/kvm_mmu.h | 2 +-
arch/arm/kvm/mmu.c | 18 +++++++++---------
arch/arm64/include/asm/kvm_mmu.h | 1 -
4 files changed, 10 insertions(+), 13 deletions(-)
diff --git a/arch/arm/include/asm/kvm_hyp.h b/arch/arm/include/asm/kvm_hyp.h
index e38fce2..6eaff28 100644
--- a/arch/arm/include/asm/kvm_hyp.h
+++ b/arch/arm/include/asm/kvm_hyp.h
@@ -25,8 +25,6 @@
#define __hyp_text __section(.hyp.text) notrace
-#define kern_hyp_va(v) (v)
-
#define __ACCESS_CP15(CRn, Op1, CRm, Op2) \
"mrc", "mcr", __stringify(p15, Op1, %0, CRn, CRm, Op2), u32
#define __ACCESS_CP15_64(Op1, CRm) \
diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h
index 73c2818..3bb803d 100644
--- a/arch/arm/include/asm/kvm_mmu.h
+++ b/arch/arm/include/asm/kvm_mmu.h
@@ -26,7 +26,7 @@
* We directly use the kernel VA for the HYP, as we can directly share
* the mapping (HTTBR "covers" TTBR1).
*/
-#define KERN_TO_HYP(kva) (kva)
+#define kern_hyp_va(kva) (kva)
/*
* KVM_MMU_CACHE_MIN_PAGES is the number of stage2 page table translation levels.
diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
index 8a0aa37..bda27b6 100644
--- a/arch/arm/kvm/mmu.c
+++ b/arch/arm/kvm/mmu.c
@@ -506,9 +506,9 @@ void free_hyp_pgds(void)
if (hyp_pgd) {
unmap_hyp_range(hyp_pgd, hyp_idmap_start, PAGE_SIZE);
for (addr = PAGE_OFFSET; virt_addr_valid(addr); addr += PGDIR_SIZE)
- unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE);
+ unmap_hyp_range(hyp_pgd, kern_hyp_va(addr), PGDIR_SIZE);
for (addr = VMALLOC_START; is_vmalloc_addr((void*)addr); addr += PGDIR_SIZE)
- unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE);
+ unmap_hyp_range(hyp_pgd, kern_hyp_va(addr), PGDIR_SIZE);
free_pages((unsigned long)hyp_pgd, hyp_pgd_order);
hyp_pgd = NULL;
@@ -670,8 +670,8 @@ int create_hyp_mappings(void *from, void *to, pgprot_t prot)
{
phys_addr_t phys_addr;
unsigned long virt_addr;
- unsigned long start = KERN_TO_HYP((unsigned long)from);
- unsigned long end = KERN_TO_HYP((unsigned long)to);
+ unsigned long start = kern_hyp_va((unsigned long)from);
+ unsigned long end = kern_hyp_va((unsigned long)to);
if (is_kernel_in_hyp_mode())
return 0;
@@ -705,8 +705,8 @@ int create_hyp_mappings(void *from, void *to, pgprot_t prot)
*/
int create_hyp_io_mappings(void *from, void *to, phys_addr_t phys_addr)
{
- unsigned long start = KERN_TO_HYP((unsigned long)from);
- unsigned long end = KERN_TO_HYP((unsigned long)to);
+ unsigned long start = kern_hyp_va((unsigned long)from);
+ unsigned long end = kern_hyp_va((unsigned long)to);
if (is_kernel_in_hyp_mode())
return 0;
@@ -1711,10 +1711,10 @@ int kvm_mmu_init(void)
kvm_info("IDMAP page: %lx\n", hyp_idmap_start);
kvm_info("HYP VA range: %lx:%lx\n",
- KERN_TO_HYP(PAGE_OFFSET), KERN_TO_HYP(~0UL));
+ kern_hyp_va(PAGE_OFFSET), kern_hyp_va(~0UL));
- if (hyp_idmap_start >= KERN_TO_HYP(PAGE_OFFSET) &&
- hyp_idmap_start < KERN_TO_HYP(~0UL)) {
+ if (hyp_idmap_start >= kern_hyp_va(PAGE_OFFSET) &&
+ hyp_idmap_start < kern_hyp_va(~0UL)) {
/*
* The idmap page is intersecting with the VA space,
* it is not safe to continue further.
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 9226f8b..b6bb834 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -133,7 +133,6 @@ static inline unsigned long __kern_hyp_va(unsigned long v)
}
#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v)))
-#define KERN_TO_HYP(v) kern_hyp_va(v)
/*
* We currently only support a 40bit IPA.
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 25/55] arm64: KVM: Clean up a condition
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (23 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 24/55] arm/arm64: Get rid of KERN_TO_HYP Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 26/55] KVM: arm/arm64: vgic: Move redistributor kvm_io_devices Marc Zyngier
` (30 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Dan Carpenter <dan.carpenter@oracle.com>
My static checker complains that this condition looks like it should be
== instead of =. This isn't a fast path, so we don't need to be fancy.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
arch/arm64/kvm/sys_regs.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index a57d650..b0b225c 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -1546,7 +1546,7 @@ static void unhandled_cp_access(struct kvm_vcpu *vcpu,
struct sys_reg_params *params)
{
u8 hsr_ec = kvm_vcpu_trap_get_class(vcpu);
- int cp;
+ int cp = -1;
switch(hsr_ec) {
case ESR_ELx_EC_CP15_32:
@@ -1558,7 +1558,7 @@ static void unhandled_cp_access(struct kvm_vcpu *vcpu,
cp = 14;
break;
default:
- WARN_ON((cp = -1));
+ WARN_ON(1);
}
kvm_err("Unsupported guest CP%d access at: %08lx\n",
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 26/55] KVM: arm/arm64: vgic: Move redistributor kvm_io_devices
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (24 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 25/55] arm64: KVM: Clean up a condition Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 27/55] KVM: arm/arm64: vgic: Check return value for kvm_register_vgic_device Marc Zyngier
` (29 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
Logically a GICv3 redistributor is assigned to a (v)CPU, so we should
aim to keep redistributor related variables out of our struct vgic_dist.
Let's start by replacing the redistributor related kvm_io_device array
with two members in our existing struct vgic_cpu, which are naturally
per-VCPU and thus don't require any allocation / freeing.
So apart from the better fit with the redistributor design this saves
some code as well.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 8 +++++++-
virt/kvm/arm/vgic/vgic-init.c | 1 -
virt/kvm/arm/vgic/vgic-mmio-v3.c | 22 ++++++++--------------
3 files changed, 15 insertions(+), 16 deletions(-)
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 1264037..5142e2a 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -145,7 +145,6 @@ struct vgic_dist {
struct vgic_irq *spis;
struct vgic_io_device dist_iodev;
- struct vgic_io_device *redist_iodevs;
};
struct vgic_v2_cpu_if {
@@ -193,6 +192,13 @@ struct vgic_cpu {
struct list_head ap_list_head;
u64 live_lrs;
+
+ /*
+ * Members below are used with GICv3 emulation only and represent
+ * parts of the redistributor.
+ */
+ struct vgic_io_device rd_iodev;
+ struct vgic_io_device sgi_iodev;
};
int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index a1442f7..90cae48 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -271,7 +271,6 @@ static void kvm_vgic_dist_destroy(struct kvm *kvm)
dist->initialized = false;
kfree(dist->spis);
- kfree(dist->redist_iodevs);
dist->nr_spis = 0;
mutex_unlock(&kvm->lock);
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index a0c515a..fc7b6c9 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -285,21 +285,14 @@ unsigned int vgic_v3_init_dist_iodev(struct vgic_io_device *dev)
int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
{
- int nr_vcpus = atomic_read(&kvm->online_vcpus);
struct kvm_vcpu *vcpu;
- struct vgic_io_device *devices;
int c, ret = 0;
- devices = kmalloc(sizeof(struct vgic_io_device) * nr_vcpus * 2,
- GFP_KERNEL);
- if (!devices)
- return -ENOMEM;
-
kvm_for_each_vcpu(c, vcpu, kvm) {
gpa_t rd_base = redist_base_address + c * SZ_64K * 2;
gpa_t sgi_base = rd_base + SZ_64K;
- struct vgic_io_device *rd_dev = &devices[c * 2];
- struct vgic_io_device *sgi_dev = &devices[c * 2 + 1];
+ struct vgic_io_device *rd_dev = &vcpu->arch.vgic_cpu.rd_iodev;
+ struct vgic_io_device *sgi_dev = &vcpu->arch.vgic_cpu.sgi_iodev;
kvm_iodevice_init(&rd_dev->dev, &kvm_io_gic_ops);
rd_dev->base_addr = rd_base;
@@ -335,14 +328,15 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
if (ret) {
/* The current c failed, so we start with the previous one. */
for (c--; c >= 0; c--) {
+ struct vgic_cpu *vgic_cpu;
+
+ vcpu = kvm_get_vcpu(kvm, c);
+ vgic_cpu = &vcpu->arch.vgic_cpu;
kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS,
- &devices[c * 2].dev);
+ &vgic_cpu->rd_iodev.dev);
kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS,
- &devices[c * 2 + 1].dev);
+ &vgic_cpu->sgi_iodev.dev);
}
- kfree(devices);
- } else {
- kvm->arch.vgic.redist_iodevs = devices;
}
return ret;
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 27/55] KVM: arm/arm64: vgic: Check return value for kvm_register_vgic_device
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (25 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 26/55] KVM: arm/arm64: vgic: Move redistributor kvm_io_devices Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 28/55] KVM: Extend struct kvm_msi to hold a 32-bit device ID Marc Zyngier
` (28 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
kvm_register_device_ops() can return an error, so lets check its return
value and propagate this up the call chain.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-kvm-device.c | 15 +++++++++------
virt/kvm/arm/vgic/vgic-v2.c | 11 ++++++++---
virt/kvm/arm/vgic/vgic-v3.c | 15 +++++++++++++--
virt/kvm/arm/vgic/vgic.h | 2 +-
4 files changed, 31 insertions(+), 12 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-kvm-device.c b/virt/kvm/arm/vgic/vgic-kvm-device.c
index 0130c4b..2f24f13 100644
--- a/virt/kvm/arm/vgic/vgic-kvm-device.c
+++ b/virt/kvm/arm/vgic/vgic-kvm-device.c
@@ -210,20 +210,24 @@ static void vgic_destroy(struct kvm_device *dev)
kfree(dev);
}
-void kvm_register_vgic_device(unsigned long type)
+int kvm_register_vgic_device(unsigned long type)
{
+ int ret = -ENODEV;
+
switch (type) {
case KVM_DEV_TYPE_ARM_VGIC_V2:
- kvm_register_device_ops(&kvm_arm_vgic_v2_ops,
- KVM_DEV_TYPE_ARM_VGIC_V2);
+ ret = kvm_register_device_ops(&kvm_arm_vgic_v2_ops,
+ KVM_DEV_TYPE_ARM_VGIC_V2);
break;
#ifdef CONFIG_KVM_ARM_VGIC_V3
case KVM_DEV_TYPE_ARM_VGIC_V3:
- kvm_register_device_ops(&kvm_arm_vgic_v3_ops,
- KVM_DEV_TYPE_ARM_VGIC_V3);
+ ret = kvm_register_device_ops(&kvm_arm_vgic_v3_ops,
+ KVM_DEV_TYPE_ARM_VGIC_V3);
break;
#endif
}
+
+ return ret;
}
/** vgic_attr_regs_access: allows user space to read/write VGIC registers
@@ -428,4 +432,3 @@ struct kvm_device_ops kvm_arm_vgic_v3_ops = {
};
#endif /* CONFIG_KVM_ARM_VGIC_V3 */
-
diff --git a/virt/kvm/arm/vgic/vgic-v2.c b/virt/kvm/arm/vgic/vgic-v2.c
index e31405e..079bf67 100644
--- a/virt/kvm/arm/vgic/vgic-v2.c
+++ b/virt/kvm/arm/vgic/vgic-v2.c
@@ -332,20 +332,25 @@ int vgic_v2_probe(const struct gic_kvm_info *info)
vtr = readl_relaxed(kvm_vgic_global_state.vctrl_base + GICH_VTR);
kvm_vgic_global_state.nr_lr = (vtr & 0x3f) + 1;
+ ret = kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+ if (ret) {
+ kvm_err("Cannot register GICv2 KVM device\n");
+ iounmap(kvm_vgic_global_state.vctrl_base);
+ return ret;
+ }
+
ret = create_hyp_io_mappings(kvm_vgic_global_state.vctrl_base,
kvm_vgic_global_state.vctrl_base +
resource_size(&info->vctrl),
info->vctrl.start);
-
if (ret) {
kvm_err("Cannot map VCTRL into hyp\n");
+ kvm_unregister_device_ops(KVM_DEV_TYPE_ARM_VGIC_V2);
iounmap(kvm_vgic_global_state.vctrl_base);
return ret;
}
kvm_vgic_global_state.can_emulate_gicv2 = true;
- kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
-
kvm_vgic_global_state.vcpu_base = info->vcpu.start;
kvm_vgic_global_state.type = VGIC_V2;
kvm_vgic_global_state.max_gic_vcpus = VGIC_V2_MAX_CPUS;
diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c
index 346b4ad..e48a22e 100644
--- a/virt/kvm/arm/vgic/vgic-v3.c
+++ b/virt/kvm/arm/vgic/vgic-v3.c
@@ -296,6 +296,7 @@ out:
int vgic_v3_probe(const struct gic_kvm_info *info)
{
u32 ich_vtr_el2 = kvm_call_hyp(__vgic_v3_get_ich_vtr_el2);
+ int ret;
/*
* The ListRegs field is 5 bits, but there is a architectural
@@ -319,12 +320,22 @@ int vgic_v3_probe(const struct gic_kvm_info *info)
} else {
kvm_vgic_global_state.vcpu_base = info->vcpu.start;
kvm_vgic_global_state.can_emulate_gicv2 = true;
- kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+ ret = kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V2);
+ if (ret) {
+ kvm_err("Cannot register GICv2 KVM device.\n");
+ return ret;
+ }
kvm_info("vgic-v2@%llx\n", info->vcpu.start);
}
+ ret = kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V3);
+ if (ret) {
+ kvm_err("Cannot register GICv3 KVM device.\n");
+ kvm_unregister_device_ops(KVM_DEV_TYPE_ARM_VGIC_V2);
+ return ret;
+ }
+
if (kvm_vgic_global_state.vcpu_base == 0)
kvm_info("disabling GICv2 emulation\n");
- kvm_register_vgic_device(KVM_DEV_TYPE_ARM_VGIC_V3);
kvm_vgic_global_state.vctrl_base = NULL;
kvm_vgic_global_state.type = VGIC_V3;
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 7b300ca..c752152 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -124,7 +124,7 @@ static inline int vgic_register_redist_iodevs(struct kvm *kvm,
}
#endif
-void kvm_register_vgic_device(unsigned long type);
+int kvm_register_vgic_device(unsigned long type);
int vgic_lazy_init(struct kvm *kvm);
int vgic_init(struct kvm *kvm);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 28/55] KVM: Extend struct kvm_msi to hold a 32-bit device ID
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (26 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 27/55] KVM: arm/arm64: vgic: Check return value for kvm_register_vgic_device Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 29/55] KVM: arm/arm64: Extend arch CAP checks to allow per-VM capabilities Marc Zyngier
` (27 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
The ARM GICv3 ITS MSI controller requires a device ID to be able to
assign the proper interrupt vector. On real hardware, this ID is
sampled from the bus. To be able to emulate an ITS controller, extend
the KVM MSI interface to let userspace provide such a device ID. For
PCI devices, the device ID is simply the 16-bit bus-device-function
triplet, which should be easily available to the userland tool.
Also there is a new KVM capability which advertises whether the
current VM requires a device ID to be set along with the MSI data.
This flag is still reported as not available everywhere, later we will
enable it when ITS emulation is used.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Eric Auger <eric.auger@linaro.org>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
Documentation/virtual/kvm/api.txt | 12 ++++++++++--
include/uapi/linux/kvm.h | 5 ++++-
2 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index 09efa9e..6551311 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -2175,10 +2175,18 @@ struct kvm_msi {
__u32 address_hi;
__u32 data;
__u32 flags;
- __u8 pad[16];
+ __u32 devid;
+ __u8 pad[12];
};
-No flags are defined so far. The corresponding field must be 0.
+flags: KVM_MSI_VALID_DEVID: devid contains a valid value
+devid: If KVM_MSI_VALID_DEVID is set, contains a unique device identifier
+ for the device that wrote the MSI message.
+ For PCI, this is usually a BFD identifier in the lower 16 bits.
+
+The per-VM KVM_CAP_MSI_DEVID capability advertises the need to provide
+the device ID. If this capability is not set, userland cannot rely on
+the kernel to allow the KVM_MSI_VALID_DEVID flag being set.
4.71 KVM_CREATE_PIT2
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index 05ebf47..7de96f5 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -866,6 +866,7 @@ struct kvm_ppc_smmu_info {
#define KVM_CAP_ARM_PMU_V3 126
#define KVM_CAP_VCPU_ATTRIBUTES 127
#define KVM_CAP_MAX_VCPU_ID 128
+#define KVM_CAP_MSI_DEVID 129
#ifdef KVM_CAP_IRQ_ROUTING
@@ -1024,12 +1025,14 @@ struct kvm_one_reg {
__u64 addr;
};
+#define KVM_MSI_VALID_DEVID (1U << 0)
struct kvm_msi {
__u32 address_lo;
__u32 address_hi;
__u32 data;
__u32 flags;
- __u8 pad[16];
+ __u32 devid;
+ __u8 pad[12];
};
struct kvm_arm_device_addr {
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 29/55] KVM: arm/arm64: Extend arch CAP checks to allow per-VM capabilities
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (27 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 28/55] KVM: Extend struct kvm_msi to hold a 32-bit device ID Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 30/55] KVM: kvm_io_bus: Add kvm_io_bus_get_dev() call Marc Zyngier
` (26 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
KVM capabilities can be a per-VM property, though ARM/ARM64 currently
does not pass on the VM pointer to the architecture specific
capability handlers.
Add a "struct kvm*" parameter to those function to later allow proper
per-VM capability reporting.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Eric Auger <eric.auger@linaro.org>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
arch/arm/include/asm/kvm_host.h | 2 +-
arch/arm/kvm/arm.c | 2 +-
arch/arm64/include/asm/kvm_host.h | 2 +-
arch/arm64/kvm/reset.c | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
index 58d0b69..de338d9 100644
--- a/arch/arm/include/asm/kvm_host.h
+++ b/arch/arm/include/asm/kvm_host.h
@@ -272,7 +272,7 @@ static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr,
kvm_call_hyp((void *)virt_to_idmap(__kvm_hyp_reset), vector_ptr);
}
-static inline int kvm_arch_dev_ioctl_check_extension(long ext)
+static inline int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
{
return 0;
}
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 7cf266c..972075c 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -201,7 +201,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
r = KVM_MAX_VCPUS;
break;
default:
- r = kvm_arch_dev_ioctl_check_extension(ext);
+ r = kvm_arch_dev_ioctl_check_extension(kvm, ext);
break;
}
return r;
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 69d5cc2d..3eda975 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -47,7 +47,7 @@
int __attribute_const__ kvm_target_cpu(void);
int kvm_reset_vcpu(struct kvm_vcpu *vcpu);
-int kvm_arch_dev_ioctl_check_extension(long ext);
+int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext);
void __extended_idmap_trampoline(phys_addr_t boot_pgd, phys_addr_t idmap_start);
struct kvm_arch {
diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
index 79f3248..e95d4f6 100644
--- a/arch/arm64/kvm/reset.c
+++ b/arch/arm64/kvm/reset.c
@@ -65,7 +65,7 @@ static bool cpu_has_32bit_el1(void)
* We currently assume that the number of HW registers is uniform
* across all CPUs (see cpuinfo_sanity_check).
*/
-int kvm_arch_dev_ioctl_check_extension(long ext)
+int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
{
int r;
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 30/55] KVM: kvm_io_bus: Add kvm_io_bus_get_dev() call
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (28 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 29/55] KVM: arm/arm64: Extend arch CAP checks to allow per-VM capabilities Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 31/55] KVM: arm/arm64: vgic: Add refcounting for IRQs Marc Zyngier
` (25 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
The kvm_io_bus framework is a nice place of holding information about
various MMIO regions for kernel emulated devices.
Add a call to retrieve the kvm_io_device structure which is associated
with a certain MMIO address. This avoids to duplicate kvm_io_bus'
knowledge of MMIO regions without having to fake MMIO calls if a user
needs the device a certain MMIO address belongs to.
This will be used by the ITS emulation to get the associated ITS device
when someone triggers an MSI via an ioctl from userspace.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/linux/kvm_host.h | 2 ++
virt/kvm/kvm_main.c | 24 ++++++++++++++++++++++++
2 files changed, 26 insertions(+)
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 0640ee9..614a981 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -164,6 +164,8 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
int len, struct kvm_io_device *dev);
int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
struct kvm_io_device *dev);
+struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
+ gpa_t addr);
#ifdef CONFIG_KVM_ASYNC_PF
struct kvm_async_pf {
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index ef54b4c..bd2eb92 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -3496,6 +3496,30 @@ int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
return r;
}
+struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
+ gpa_t addr)
+{
+ struct kvm_io_bus *bus;
+ int dev_idx, srcu_idx;
+ struct kvm_io_device *iodev = NULL;
+
+ srcu_idx = srcu_read_lock(&kvm->srcu);
+
+ bus = srcu_dereference(kvm->buses[bus_idx], &kvm->srcu);
+
+ dev_idx = kvm_io_bus_get_first_dev(bus, addr, 1);
+ if (dev_idx < 0)
+ goto out_unlock;
+
+ iodev = bus->range[dev_idx].dev;
+
+out_unlock:
+ srcu_read_unlock(&kvm->srcu, srcu_idx);
+
+ return iodev;
+}
+EXPORT_SYMBOL_GPL(kvm_io_bus_get_dev);
+
static struct notifier_block kvm_cpu_notifier = {
.notifier_call = kvm_cpu_hotplug,
};
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 31/55] KVM: arm/arm64: vgic: Add refcounting for IRQs
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (29 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 30/55] KVM: kvm_io_bus: Add kvm_io_bus_get_dev() call Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 32/55] irqchip/gic-v3: Refactor and add GICv3 definitions Marc Zyngier
` (24 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
In the moment our struct vgic_irq's are statically allocated at guest
creation time. So getting a pointer to an IRQ structure is trivial and
safe. LPIs are more dynamic, they can be mapped and unmapped at any time
during the guest's _runtime_.
In preparation for supporting LPIs we introduce reference counting for
those structures using the kernel's kref infrastructure.
Since private IRQs and SPIs are statically allocated, we avoid actually
refcounting them, since they would never be released anyway.
But we take provisions to increase the refcount when an IRQ gets onto a
VCPU list and decrease it when it gets removed. Also this introduces
vgic_put_irq(), which wraps kref_put and hides the release function from
the callers.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 1 +
virt/kvm/arm/vgic/vgic-init.c | 2 ++
virt/kvm/arm/vgic/vgic-mmio-v2.c | 8 +++++++
virt/kvm/arm/vgic/vgic-mmio-v3.c | 20 ++++++++++------
virt/kvm/arm/vgic/vgic-mmio.c | 25 ++++++++++++++++++-
virt/kvm/arm/vgic/vgic-v2.c | 1 +
virt/kvm/arm/vgic/vgic-v3.c | 1 +
virt/kvm/arm/vgic/vgic.c | 52 ++++++++++++++++++++++++++++++++++++----
virt/kvm/arm/vgic/vgic.h | 1 +
9 files changed, 99 insertions(+), 12 deletions(-)
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 5142e2a..450b4da 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -96,6 +96,7 @@ struct vgic_irq {
bool active; /* not used for LPIs */
bool enabled;
bool hw; /* Tied to HW IRQ */
+ struct kref refcount; /* Used for LPIs */
u32 hwintid; /* HW INTID number */
union {
u8 targets; /* GICv2 target VCPUs mask */
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 90cae48..ac3c1a5 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -177,6 +177,7 @@ static int kvm_vgic_dist_init(struct kvm *kvm, unsigned int nr_spis)
spin_lock_init(&irq->irq_lock);
irq->vcpu = NULL;
irq->target_vcpu = vcpu0;
+ kref_init(&irq->refcount);
if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V2)
irq->targets = 0;
else
@@ -211,6 +212,7 @@ static void kvm_vgic_vcpu_init(struct kvm_vcpu *vcpu)
irq->vcpu = NULL;
irq->target_vcpu = vcpu;
irq->targets = 1U << vcpu->vcpu_id;
+ kref_init(&irq->refcount);
if (vgic_irq_is_sgi(i)) {
/* SGIs */
irq->enabled = 1;
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v2.c b/virt/kvm/arm/vgic/vgic-mmio-v2.c
index a213936..4152348 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v2.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v2.c
@@ -102,6 +102,7 @@ static void vgic_mmio_write_sgir(struct kvm_vcpu *source_vcpu,
irq->source |= 1U << source_vcpu->vcpu_id;
vgic_queue_irq_unlock(source_vcpu->kvm, irq);
+ vgic_put_irq(source_vcpu->kvm, irq);
}
}
@@ -116,6 +117,8 @@ static unsigned long vgic_mmio_read_target(struct kvm_vcpu *vcpu,
struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
val |= (u64)irq->targets << (i * 8);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return val;
@@ -143,6 +146,7 @@ static void vgic_mmio_write_target(struct kvm_vcpu *vcpu,
irq->target_vcpu = kvm_get_vcpu(vcpu->kvm, target);
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -157,6 +161,8 @@ static unsigned long vgic_mmio_read_sgipend(struct kvm_vcpu *vcpu,
struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
val |= (u64)irq->source << (i * 8);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return val;
}
@@ -178,6 +184,7 @@ static void vgic_mmio_write_sgipendc(struct kvm_vcpu *vcpu,
irq->pending = false;
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -201,6 +208,7 @@ static void vgic_mmio_write_sgipends(struct kvm_vcpu *vcpu,
} else {
spin_unlock(&irq->irq_lock);
}
+ vgic_put_irq(vcpu->kvm, irq);
}
}
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index fc7b6c9..bfcafbd 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -80,15 +80,17 @@ static unsigned long vgic_mmio_read_irouter(struct kvm_vcpu *vcpu,
{
int intid = VGIC_ADDR_TO_INTID(addr, 64);
struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, NULL, intid);
+ unsigned long ret = 0;
if (!irq)
return 0;
/* The upper word is RAZ for us. */
- if (addr & 4)
- return 0;
+ if (!(addr & 4))
+ ret = extract_bytes(READ_ONCE(irq->mpidr), addr & 7, len);
- return extract_bytes(READ_ONCE(irq->mpidr), addr & 7, len);
+ vgic_put_irq(vcpu->kvm, irq);
+ return ret;
}
static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
@@ -96,15 +98,17 @@ static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
unsigned long val)
{
int intid = VGIC_ADDR_TO_INTID(addr, 64);
- struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, NULL, intid);
-
- if (!irq)
- return;
+ struct vgic_irq *irq;
/* The upper word is WI for us since we don't implement Aff3. */
if (addr & 4)
return;
+ irq = vgic_get_irq(vcpu->kvm, NULL, intid);
+
+ if (!irq)
+ return;
+
spin_lock(&irq->irq_lock);
/* We only care about and preserve Aff0, Aff1 and Aff2. */
@@ -112,6 +116,7 @@ static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
irq->target_vcpu = kvm_mpidr_to_vcpu(vcpu->kvm, irq->mpidr);
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
static unsigned long vgic_mmio_read_v3r_typer(struct kvm_vcpu *vcpu,
@@ -445,5 +450,6 @@ void vgic_v3_dispatch_sgi(struct kvm_vcpu *vcpu, u64 reg)
irq->pending = true;
vgic_queue_irq_unlock(vcpu->kvm, irq);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
diff --git a/virt/kvm/arm/vgic/vgic-mmio.c b/virt/kvm/arm/vgic/vgic-mmio.c
index 9f6fab7..5e79e01 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.c
+++ b/virt/kvm/arm/vgic/vgic-mmio.c
@@ -56,6 +56,8 @@ unsigned long vgic_mmio_read_enable(struct kvm_vcpu *vcpu,
if (irq->enabled)
value |= (1U << i);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return value;
@@ -74,6 +76,8 @@ void vgic_mmio_write_senable(struct kvm_vcpu *vcpu,
spin_lock(&irq->irq_lock);
irq->enabled = true;
vgic_queue_irq_unlock(vcpu->kvm, irq);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -92,6 +96,7 @@ void vgic_mmio_write_cenable(struct kvm_vcpu *vcpu,
irq->enabled = false;
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -108,6 +113,8 @@ unsigned long vgic_mmio_read_pending(struct kvm_vcpu *vcpu,
if (irq->pending)
value |= (1U << i);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return value;
@@ -129,6 +136,7 @@ void vgic_mmio_write_spending(struct kvm_vcpu *vcpu,
irq->soft_pending = true;
vgic_queue_irq_unlock(vcpu->kvm, irq);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -152,6 +160,7 @@ void vgic_mmio_write_cpending(struct kvm_vcpu *vcpu,
}
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -168,6 +177,8 @@ unsigned long vgic_mmio_read_active(struct kvm_vcpu *vcpu,
if (irq->active)
value |= (1U << i);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return value;
@@ -242,6 +253,7 @@ void vgic_mmio_write_cactive(struct kvm_vcpu *vcpu,
for_each_set_bit(i, &val, len * 8) {
struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
vgic_mmio_change_active(vcpu, irq, false);
+ vgic_put_irq(vcpu->kvm, irq);
}
vgic_change_active_finish(vcpu, intid);
}
@@ -257,6 +269,7 @@ void vgic_mmio_write_sactive(struct kvm_vcpu *vcpu,
for_each_set_bit(i, &val, len * 8) {
struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
vgic_mmio_change_active(vcpu, irq, true);
+ vgic_put_irq(vcpu->kvm, irq);
}
vgic_change_active_finish(vcpu, intid);
}
@@ -272,6 +285,8 @@ unsigned long vgic_mmio_read_priority(struct kvm_vcpu *vcpu,
struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
val |= (u64)irq->priority << (i * 8);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return val;
@@ -298,6 +313,8 @@ void vgic_mmio_write_priority(struct kvm_vcpu *vcpu,
/* Narrow the priority range to what we actually support */
irq->priority = (val >> (i * 8)) & GENMASK(7, 8 - VGIC_PRI_BITS);
spin_unlock(&irq->irq_lock);
+
+ vgic_put_irq(vcpu->kvm, irq);
}
}
@@ -313,6 +330,8 @@ unsigned long vgic_mmio_read_config(struct kvm_vcpu *vcpu,
if (irq->config == VGIC_CONFIG_EDGE)
value |= (2U << (i * 2));
+
+ vgic_put_irq(vcpu->kvm, irq);
}
return value;
@@ -326,7 +345,7 @@ void vgic_mmio_write_config(struct kvm_vcpu *vcpu,
int i;
for (i = 0; i < len * 4; i++) {
- struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
+ struct vgic_irq *irq;
/*
* The configuration cannot be changed for SGIs in general,
@@ -337,14 +356,18 @@ void vgic_mmio_write_config(struct kvm_vcpu *vcpu,
if (intid + i < VGIC_NR_PRIVATE_IRQS)
continue;
+ irq = vgic_get_irq(vcpu->kvm, vcpu, intid + i);
spin_lock(&irq->irq_lock);
+
if (test_bit(i * 2 + 1, &val)) {
irq->config = VGIC_CONFIG_EDGE;
} else {
irq->config = VGIC_CONFIG_LEVEL;
irq->pending = irq->line_level | irq->soft_pending;
}
+
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
diff --git a/virt/kvm/arm/vgic/vgic-v2.c b/virt/kvm/arm/vgic/vgic-v2.c
index 079bf67..0bf6709 100644
--- a/virt/kvm/arm/vgic/vgic-v2.c
+++ b/virt/kvm/arm/vgic/vgic-v2.c
@@ -124,6 +124,7 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
}
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c
index e48a22e..f0ac064 100644
--- a/virt/kvm/arm/vgic/vgic-v3.c
+++ b/virt/kvm/arm/vgic/vgic-v3.c
@@ -113,6 +113,7 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
}
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
}
}
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index 69b61ab..fb19a55 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -64,6 +64,28 @@ struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
return NULL;
}
+static void vgic_get_irq_kref(struct vgic_irq *irq)
+{
+ if (irq->intid < VGIC_MIN_LPI)
+ return;
+
+ kref_get(&irq->refcount);
+}
+
+/* The refcount should never drop to 0 at the moment. */
+static void vgic_irq_release(struct kref *ref)
+{
+ WARN_ON(1);
+}
+
+void vgic_put_irq(struct kvm *kvm, struct vgic_irq *irq)
+{
+ if (irq->intid < VGIC_MIN_LPI)
+ return;
+
+ kref_put(&irq->refcount, vgic_irq_release);
+}
+
/**
* kvm_vgic_target_oracle - compute the target vcpu for an irq
*
@@ -236,6 +258,11 @@ retry:
goto retry;
}
+ /*
+ * Grab a reference to the irq to reflect the fact that it is
+ * now in the ap_list.
+ */
+ vgic_get_irq_kref(irq);
list_add_tail(&irq->ap_list, &vcpu->arch.vgic_cpu.ap_list_head);
irq->vcpu = vcpu;
@@ -269,14 +296,17 @@ static int vgic_update_irq_pending(struct kvm *kvm, int cpuid,
if (!irq)
return -EINVAL;
- if (irq->hw != mapped_irq)
+ if (irq->hw != mapped_irq) {
+ vgic_put_irq(kvm, irq);
return -EINVAL;
+ }
spin_lock(&irq->irq_lock);
if (!vgic_validate_injection(irq, level)) {
/* Nothing to see here, move along... */
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(kvm, irq);
return 0;
}
@@ -288,6 +318,7 @@ static int vgic_update_irq_pending(struct kvm *kvm, int cpuid,
}
vgic_queue_irq_unlock(kvm, irq);
+ vgic_put_irq(kvm, irq);
return 0;
}
@@ -330,25 +361,28 @@ int kvm_vgic_map_phys_irq(struct kvm_vcpu *vcpu, u32 virt_irq, u32 phys_irq)
irq->hwintid = phys_irq;
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
return 0;
}
int kvm_vgic_unmap_phys_irq(struct kvm_vcpu *vcpu, unsigned int virt_irq)
{
- struct vgic_irq *irq = vgic_get_irq(vcpu->kvm, vcpu, virt_irq);
-
- BUG_ON(!irq);
+ struct vgic_irq *irq;
if (!vgic_initialized(vcpu->kvm))
return -EAGAIN;
+ irq = vgic_get_irq(vcpu->kvm, vcpu, virt_irq);
+ BUG_ON(!irq);
+
spin_lock(&irq->irq_lock);
irq->hw = false;
irq->hwintid = 0;
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
return 0;
}
@@ -386,6 +420,15 @@ retry:
list_del(&irq->ap_list);
irq->vcpu = NULL;
spin_unlock(&irq->irq_lock);
+
+ /*
+ * This vgic_put_irq call matches the
+ * vgic_get_irq_kref in vgic_queue_irq_unlock,
+ * where we added the LPI to the ap_list. As
+ * we remove the irq from the list, we drop
+ * also drop the refcount.
+ */
+ vgic_put_irq(vcpu->kvm, irq);
continue;
}
@@ -614,6 +657,7 @@ bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq)
spin_lock(&irq->irq_lock);
map_is_active = irq->hw && irq->active;
spin_unlock(&irq->irq_lock);
+ vgic_put_irq(vcpu->kvm, irq);
return map_is_active;
}
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index c752152..5b79c34 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -38,6 +38,7 @@ struct vgic_vmcr {
struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
u32 intid);
+void vgic_put_irq(struct kvm *kvm, struct vgic_irq *irq);
bool vgic_queue_irq_unlock(struct kvm *kvm, struct vgic_irq *irq);
void vgic_kick_vcpus(struct kvm *kvm);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 32/55] irqchip/gic-v3: Refactor and add GICv3 definitions
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (30 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 31/55] KVM: arm/arm64: vgic: Add refcounting for IRQs Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers Marc Zyngier
` (23 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
arm-gic-v3.h contains bit and register definitions for the GICv3 and ITS,
at least for the bits the we currently care about.
The ITS emulation needs more definitions, so add them and refactor
the memory attribute #defines to be more universally usable.
To avoid changing all users, we still provide some of the old definitons
defined with the help of the new macros.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/linux/irqchip/arm-gic-v3.h | 180 ++++++++++++++++++++++++-------------
1 file changed, 120 insertions(+), 60 deletions(-)
diff --git a/include/linux/irqchip/arm-gic-v3.h b/include/linux/irqchip/arm-gic-v3.h
index bfbd707..9442be7 100644
--- a/include/linux/irqchip/arm-gic-v3.h
+++ b/include/linux/irqchip/arm-gic-v3.h
@@ -112,34 +112,60 @@
#define GICR_WAKER_ProcessorSleep (1U << 1)
#define GICR_WAKER_ChildrenAsleep (1U << 2)
-#define GICR_PROPBASER_NonShareable (0U << 10)
-#define GICR_PROPBASER_InnerShareable (1U << 10)
-#define GICR_PROPBASER_OuterShareable (2U << 10)
-#define GICR_PROPBASER_SHAREABILITY_MASK (3UL << 10)
-#define GICR_PROPBASER_nCnB (0U << 7)
-#define GICR_PROPBASER_nC (1U << 7)
-#define GICR_PROPBASER_RaWt (2U << 7)
-#define GICR_PROPBASER_RaWb (3U << 7)
-#define GICR_PROPBASER_WaWt (4U << 7)
-#define GICR_PROPBASER_WaWb (5U << 7)
-#define GICR_PROPBASER_RaWaWt (6U << 7)
-#define GICR_PROPBASER_RaWaWb (7U << 7)
-#define GICR_PROPBASER_CACHEABILITY_MASK (7U << 7)
-#define GICR_PROPBASER_IDBITS_MASK (0x1f)
-
-#define GICR_PENDBASER_NonShareable (0U << 10)
-#define GICR_PENDBASER_InnerShareable (1U << 10)
-#define GICR_PENDBASER_OuterShareable (2U << 10)
-#define GICR_PENDBASER_SHAREABILITY_MASK (3UL << 10)
-#define GICR_PENDBASER_nCnB (0U << 7)
-#define GICR_PENDBASER_nC (1U << 7)
-#define GICR_PENDBASER_RaWt (2U << 7)
-#define GICR_PENDBASER_RaWb (3U << 7)
-#define GICR_PENDBASER_WaWt (4U << 7)
-#define GICR_PENDBASER_WaWb (5U << 7)
-#define GICR_PENDBASER_RaWaWt (6U << 7)
-#define GICR_PENDBASER_RaWaWb (7U << 7)
-#define GICR_PENDBASER_CACHEABILITY_MASK (7U << 7)
+#define GIC_BASER_CACHE_nCnB 0ULL
+#define GIC_BASER_CACHE_SameAsInner 0ULL
+#define GIC_BASER_CACHE_nC 1ULL
+#define GIC_BASER_CACHE_RaWt 2ULL
+#define GIC_BASER_CACHE_RaWb 3ULL
+#define GIC_BASER_CACHE_WaWt 4ULL
+#define GIC_BASER_CACHE_WaWb 5ULL
+#define GIC_BASER_CACHE_RaWaWt 6ULL
+#define GIC_BASER_CACHE_RaWaWb 7ULL
+#define GIC_BASER_CACHE_MASK 7ULL
+#define GIC_BASER_NonShareable 0ULL
+#define GIC_BASER_InnerShareable 1ULL
+#define GIC_BASER_OuterShareable 2ULL
+#define GIC_BASER_SHAREABILITY_MASK 3ULL
+
+#define GIC_BASER_CACHEABILITY(reg, inner_outer, type) \
+ (GIC_BASER_CACHE_##type << reg##_##inner_outer##_CACHEABILITY_SHIFT)
+
+#define GIC_BASER_SHAREABILITY(reg, type) \
+ (GIC_BASER_##type << reg##_SHAREABILITY_SHIFT)
+
+#define GICR_PROPBASER_SHAREABILITY_SHIFT (10)
+#define GICR_PROPBASER_INNER_CACHEABILITY_SHIFT (7)
+#define GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT (56)
+#define GICR_PROPBASER_SHAREABILITY_MASK \
+ GIC_BASER_SHAREABILITY(GICR_PROPBASER, SHAREABILITY_MASK)
+#define GICR_PROPBASER_INNER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, MASK)
+#define GICR_PROPBASER_OUTER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GICR_PROPBASER, OUTER, MASK)
+#define GICR_PROPBASER_CACHEABILITY_MASK GICR_PROPBASER_INNER_CACHEABILITY_MASK
+
+#define GICR_PROPBASER_InnerShareable \
+ GIC_BASER_SHAREABILITY(GICR_PROPBASER, InnerShareable)
+#define GICR_PROPBASER_nC GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, nC)
+#define GICR_PROPBASER_WaWb GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, WaWb)
+#define GICR_PROPBASER_IDBITS_MASK (0x1f)
+
+#define GICR_PENDBASER_SHAREABILITY_SHIFT (10)
+#define GICR_PENDBASER_INNER_CACHEABILITY_SHIFT (7)
+#define GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT (56)
+#define GICR_PENDBASER_SHAREABILITY_MASK \
+ GIC_BASER_SHAREABILITY(GICR_PENDBASER, SHAREABILITY_MASK)
+#define GICR_PENDBASER_INNER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, MASK)
+#define GICR_PENDBASER_OUTER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GICR_PENDBASER, OUTER, MASK)
+#define GICR_PENDBASER_CACHEABILITY_MASK GICR_PENDBASER_INNER_CACHEABILITY_MASK
+
+#define GICR_PENDBASER_InnerShareable \
+ GIC_BASER_SHAREABILITY(GICR_PENDBASER, InnerShareable)
+#define GICR_PENDBASER_nC GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, nC)
+#define GICR_PENDBASER_WaWb GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, WaWb)
+#define GICR_PENDBASER_PTZ BIT_ULL(62)
/*
* Re-Distributor registers, offsets from SGI_base
@@ -175,59 +201,74 @@
#define GITS_CWRITER 0x0088
#define GITS_CREADR 0x0090
#define GITS_BASER 0x0100
+#define GITS_IDREGS_BASE 0xffd0
+#define GITS_PIDR0 0xffe0
+#define GITS_PIDR1 0xffe4
#define GITS_PIDR2 GICR_PIDR2
+#define GITS_PIDR4 0xffd0
+#define GITS_CIDR0 0xfff0
+#define GITS_CIDR1 0xfff4
+#define GITS_CIDR2 0xfff8
+#define GITS_CIDR3 0xfffc
#define GITS_TRANSLATER 0x10040
#define GITS_CTLR_ENABLE (1U << 0)
#define GITS_CTLR_QUIESCENT (1U << 31)
+#define GITS_TYPER_PLPIS (1UL << 0)
+#define GITS_TYPER_IDBITS_SHIFT 8
#define GITS_TYPER_DEVBITS_SHIFT 13
#define GITS_TYPER_DEVBITS(r) ((((r) >> GITS_TYPER_DEVBITS_SHIFT) & 0x1f) + 1)
#define GITS_TYPER_PTA (1UL << 19)
-
-#define GITS_CBASER_VALID (1UL << 63)
-#define GITS_CBASER_nCnB (0UL << 59)
-#define GITS_CBASER_nC (1UL << 59)
-#define GITS_CBASER_RaWt (2UL << 59)
-#define GITS_CBASER_RaWb (3UL << 59)
-#define GITS_CBASER_WaWt (4UL << 59)
-#define GITS_CBASER_WaWb (5UL << 59)
-#define GITS_CBASER_RaWaWt (6UL << 59)
-#define GITS_CBASER_RaWaWb (7UL << 59)
-#define GITS_CBASER_CACHEABILITY_MASK (7UL << 59)
-#define GITS_CBASER_NonShareable (0UL << 10)
-#define GITS_CBASER_InnerShareable (1UL << 10)
-#define GITS_CBASER_OuterShareable (2UL << 10)
-#define GITS_CBASER_SHAREABILITY_MASK (3UL << 10)
+#define GITS_TYPER_HWCOLLCNT_SHIFT 24
+
+#define GITS_CBASER_VALID (1UL << 63)
+#define GITS_CBASER_SHAREABILITY_SHIFT (10)
+#define GITS_CBASER_INNER_CACHEABILITY_SHIFT (59)
+#define GITS_CBASER_OUTER_CACHEABILITY_SHIFT (53)
+#define GITS_CBASER_SHAREABILITY_MASK \
+ GIC_BASER_SHAREABILITY(GITS_CBASER, SHAREABILITY_MASK)
+#define GITS_CBASER_INNER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, MASK)
+#define GITS_CBASER_OUTER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GITS_CBASER, OUTER, MASK)
+#define GITS_CBASER_CACHEABILITY_MASK GITS_CBASER_INNER_CACHEABILITY_MASK
+
+#define GITS_CBASER_InnerShareable \
+ GIC_BASER_SHAREABILITY(GITS_CBASER, InnerShareable)
+#define GITS_CBASER_nC GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, nC)
+#define GITS_CBASER_WaWb GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, WaWb)
#define GITS_BASER_NR_REGS 8
-#define GITS_BASER_VALID (1UL << 63)
-#define GITS_BASER_nCnB (0UL << 59)
-#define GITS_BASER_nC (1UL << 59)
-#define GITS_BASER_RaWt (2UL << 59)
-#define GITS_BASER_RaWb (3UL << 59)
-#define GITS_BASER_WaWt (4UL << 59)
-#define GITS_BASER_WaWb (5UL << 59)
-#define GITS_BASER_RaWaWt (6UL << 59)
-#define GITS_BASER_RaWaWb (7UL << 59)
-#define GITS_BASER_CACHEABILITY_MASK (7UL << 59)
-#define GITS_BASER_TYPE_SHIFT (56)
+#define GITS_BASER_VALID (1UL << 63)
+#define GITS_BASER_INDIRECT (1ULL << 62)
+#define GITS_BASER_INNER_CACHEABILITY_SHIFT (59)
+#define GITS_BASER_OUTER_CACHEABILITY_SHIFT (53)
+#define GITS_BASER_INNER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GITS_BASER, INNER, MASK)
+#define GITS_BASER_OUTER_CACHEABILITY_MASK \
+ GIC_BASER_CACHEABILITY(GITS_BASER, OUTER, MASK)
+#define GITS_BASER_SHAREABILITY_MASK \
+ GIC_BASER_SHAREABILITY(GITS_BASER, SHAREABILITY_MASK)
+
+#define GITS_BASER_nC GIC_BASER_CACHEABILITY(GITS_BASER, INNER, nC)
+#define GITS_BASER_WaWb GIC_BASER_CACHEABILITY(GITS_BASER, INNER, WaWb)
+#define GITS_BASER_TYPE_SHIFT (56)
#define GITS_BASER_TYPE(r) (((r) >> GITS_BASER_TYPE_SHIFT) & 7)
-#define GITS_BASER_ENTRY_SIZE_SHIFT (48)
+#define GITS_BASER_ENTRY_SIZE_SHIFT (48)
#define GITS_BASER_ENTRY_SIZE(r) ((((r) >> GITS_BASER_ENTRY_SIZE_SHIFT) & 0xff) + 1)
-#define GITS_BASER_NonShareable (0UL << 10)
-#define GITS_BASER_InnerShareable (1UL << 10)
-#define GITS_BASER_OuterShareable (2UL << 10)
#define GITS_BASER_SHAREABILITY_SHIFT (10)
-#define GITS_BASER_SHAREABILITY_MASK (3UL << GITS_BASER_SHAREABILITY_SHIFT)
+#define GITS_BASER_InnerShareable \
+ GIC_BASER_SHAREABILITY(GITS_BASER, InnerShareable)
#define GITS_BASER_PAGE_SIZE_SHIFT (8)
#define GITS_BASER_PAGE_SIZE_4K (0UL << GITS_BASER_PAGE_SIZE_SHIFT)
#define GITS_BASER_PAGE_SIZE_16K (1UL << GITS_BASER_PAGE_SIZE_SHIFT)
#define GITS_BASER_PAGE_SIZE_64K (2UL << GITS_BASER_PAGE_SIZE_SHIFT)
#define GITS_BASER_PAGE_SIZE_MASK (3UL << GITS_BASER_PAGE_SIZE_SHIFT)
#define GITS_BASER_PAGES_MAX 256
+#define GITS_BASER_NR_PAGES(r) (((r) & 0xff) + 1)
#define GITS_BASER_TYPE_NONE 0
#define GITS_BASER_TYPE_DEVICE 1
@@ -243,7 +284,10 @@
*/
#define GITS_CMD_MAPD 0x08
#define GITS_CMD_MAPC 0x09
-#define GITS_CMD_MAPVI 0x0a
+#define GITS_CMD_MAPTI 0x0a
+/* older GIC documentation used MAPVI for this command */
+#define GITS_CMD_MAPVI GITS_CMD_MAPTI
+#define GITS_CMD_MAPI 0x0b
#define GITS_CMD_MOVI 0x01
#define GITS_CMD_DISCARD 0x0f
#define GITS_CMD_INV 0x0c
@@ -254,6 +298,22 @@
#define GITS_CMD_SYNC 0x05
/*
+ * ITS error numbers
+ */
+#define E_ITS_MOVI_UNMAPPED_INTERRUPT 0x010107
+#define E_ITS_MOVI_UNMAPPED_COLLECTION 0x010109
+#define E_ITS_CLEAR_UNMAPPED_INTERRUPT 0x010507
+#define E_ITS_MAPD_DEVICE_OOR 0x010801
+#define E_ITS_MAPC_PROCNUM_OOR 0x010902
+#define E_ITS_MAPC_COLLECTION_OOR 0x010903
+#define E_ITS_MAPTI_UNMAPPED_DEVICE 0x010a04
+#define E_ITS_MAPTI_PHYSICALID_OOR 0x010a06
+#define E_ITS_INV_UNMAPPED_INTERRUPT 0x010c07
+#define E_ITS_INVALL_UNMAPPED_COLLECTION 0x010d09
+#define E_ITS_MOVALL_PROCNUM_OOR 0x010e01
+#define E_ITS_DISCARD_UNMAPPED_INTERRUPT 0x010f07
+
+/*
* CPU interface registers
*/
#define ICC_CTLR_EL1_EOImode_drop_dir (0U << 1)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (31 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 32/55] irqchip/gic-v3: Refactor and add GICv3 definitions Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-08-01 18:20 ` Christoffer Dall
2016-07-22 17:28 ` [PATCH 34/55] KVM: arm64: vgic-its: Introduce ITS emulation file with MMIO framework Marc Zyngier
` (22 subsequent siblings)
55 siblings, 1 reply; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
In the GICv3 redistributor there are the PENDBASER and PROPBASER
registers which we did not emulate so far, as they only make sense
when having an ITS. In preparation for that emulate those MMIO
accesses by storing the 64-bit data written into it into a variable
which we later read in the ITS emulation.
We also sanitise the registers, making sure RES0 regions are respected
and checking for valid memory attributes.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 13 ++++
virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
4 files changed, 181 insertions(+), 4 deletions(-)
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 450b4da..df2dec5 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -146,6 +146,14 @@ struct vgic_dist {
struct vgic_irq *spis;
struct vgic_io_device dist_iodev;
+
+ /*
+ * Contains the attributes and gpa of the LPI configuration table.
+ * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
+ * one address across all redistributors.
+ * GICv3 spec: 6.1.2 "LPI Configuration tables"
+ */
+ u64 propbaser;
};
struct vgic_v2_cpu_if {
@@ -200,6 +208,11 @@ struct vgic_cpu {
*/
struct vgic_io_device rd_iodev;
struct vgic_io_device sgi_iodev;
+
+ /* Contains the attributes and gpa of the LPI pending tables. */
+ u64 pendbaser;
+
+ bool lpis_enabled;
};
int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index bfcafbd..278bfbb 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
}
+/* allows updates of any half of a 64-bit register (or the whole thing) */
+static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
+ unsigned long val)
+{
+ int lower = (offset & 4) * 8;
+ int upper = lower + 8 * len - 1;
+
+ reg &= ~GENMASK_ULL(upper, lower);
+ val &= GENMASK_ULL(len * 8 - 1, 0);
+
+ return reg | ((u64)val << lower);
+}
+
static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
gpa_t addr, unsigned int len)
{
@@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
return 0;
}
+/* We want to avoid outer shareable. */
+u64 vgic_sanitise_shareability(u64 field)
+{
+ switch (field) {
+ case GIC_BASER_OuterShareable:
+ return GIC_BASER_InnerShareable;
+ default:
+ return field;
+ }
+}
+
+/* Avoid any inner non-cacheable mapping. */
+u64 vgic_sanitise_inner_cacheability(u64 field)
+{
+ switch (field) {
+ case GIC_BASER_CACHE_nCnB:
+ case GIC_BASER_CACHE_nC:
+ return GIC_BASER_CACHE_RaWb;
+ default:
+ return field;
+ }
+}
+
+/* Non-cacheable or same-as-inner are OK. */
+u64 vgic_sanitise_outer_cacheability(u64 field)
+{
+ switch (field) {
+ case GIC_BASER_CACHE_SameAsInner:
+ case GIC_BASER_CACHE_nC:
+ return field;
+ default:
+ return GIC_BASER_CACHE_nC;
+ }
+}
+
+u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
+ u64 (*sanitise_fn)(u64))
+{
+ u64 field = (reg & field_mask) >> field_shift;
+
+ field = sanitise_fn(field) << field_shift;
+ return (reg & ~field_mask) | field;
+}
+
+#define PROPBASER_RES0_MASK \
+ (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
+#define PENDBASER_RES0_MASK \
+ (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
+ GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
+
+static u64 vgic_sanitise_pendbaser(u64 reg)
+{
+ reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
+ GICR_PENDBASER_SHAREABILITY_SHIFT,
+ vgic_sanitise_shareability);
+ reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
+ GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
+ vgic_sanitise_inner_cacheability);
+ reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
+ GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
+ vgic_sanitise_outer_cacheability);
+
+ reg &= ~PENDBASER_RES0_MASK;
+ reg &= ~GENMASK_ULL(51, 48);
+
+ return reg;
+}
+
+static u64 vgic_sanitise_propbaser(u64 reg)
+{
+ reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
+ GICR_PROPBASER_SHAREABILITY_SHIFT,
+ vgic_sanitise_shareability);
+ reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
+ GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
+ vgic_sanitise_inner_cacheability);
+ reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
+ GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
+ vgic_sanitise_outer_cacheability);
+
+ reg &= ~PROPBASER_RES0_MASK;
+ reg &= ~GENMASK_ULL(51, 48);
+ return reg;
+}
+
+static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
+ gpa_t addr, unsigned int len)
+{
+ struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
+
+ return extract_bytes(dist->propbaser, addr & 7, len);
+}
+
+static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
+ struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+ u64 propbaser = dist->propbaser;
+
+ /* Storing a value with LPIs already enabled is undefined */
+ if (vgic_cpu->lpis_enabled)
+ return;
+
+ propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
+ propbaser = vgic_sanitise_propbaser(propbaser);
+
+ dist->propbaser = propbaser;
+}
+
+static unsigned long vgic_mmio_read_pendbase(struct kvm_vcpu *vcpu,
+ gpa_t addr, unsigned int len)
+{
+ struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+
+ return extract_bytes(vgic_cpu->pendbaser, addr & 7, len);
+}
+
+static void vgic_mmio_write_pendbase(struct kvm_vcpu *vcpu,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+ u64 pendbaser = vgic_cpu->pendbaser;
+
+ /* Storing a value with LPIs already enabled is undefined */
+ if (vgic_cpu->lpis_enabled)
+ return;
+
+ pendbaser = update_64bit_reg(pendbaser, addr & 4, len, val);
+ pendbaser = vgic_sanitise_pendbaser(pendbaser);
+
+ vgic_cpu->pendbaser = pendbaser;
+}
+
/*
* The GICv3 per-IRQ registers are split to control PPIs and SGIs in the
* redistributors, while SPIs are covered by registers in the distributor
@@ -232,10 +381,10 @@ static const struct vgic_register_region vgic_v3_rdbase_registers[] = {
vgic_mmio_read_v3r_typer, vgic_mmio_write_wi, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_DESC_WITH_LENGTH(GICR_PROPBASER,
- vgic_mmio_read_raz, vgic_mmio_write_wi, 8,
+ vgic_mmio_read_propbase, vgic_mmio_write_propbase, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_DESC_WITH_LENGTH(GICR_PENDBASER,
- vgic_mmio_read_raz, vgic_mmio_write_wi, 8,
+ vgic_mmio_read_pendbase, vgic_mmio_write_pendbase, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_DESC_WITH_LENGTH(GICR_IDREGS,
vgic_mmio_read_v3_idregs, vgic_mmio_write_wi, 48,
diff --git a/virt/kvm/arm/vgic/vgic-mmio.h b/virt/kvm/arm/vgic/vgic-mmio.h
index 8509014..71aa39d 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.h
+++ b/virt/kvm/arm/vgic/vgic-mmio.h
@@ -147,4 +147,12 @@ unsigned int vgic_v2_init_dist_iodev(struct vgic_io_device *dev);
unsigned int vgic_v3_init_dist_iodev(struct vgic_io_device *dev);
+#ifdef CONFIG_KVM_ARM_VGIC_V3
+u64 vgic_sanitise_outer_cacheability(u64 reg);
+u64 vgic_sanitise_inner_cacheability(u64 reg);
+u64 vgic_sanitise_shareability(u64 reg);
+u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
+ u64 (*sanitise_fn)(u64));
+#endif
+
#endif
diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c
index f0ac064..6f8f31f 100644
--- a/virt/kvm/arm/vgic/vgic-v3.c
+++ b/virt/kvm/arm/vgic/vgic-v3.c
@@ -191,6 +191,11 @@ void vgic_v3_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
vmcrp->pmr = (vmcr & ICH_VMCR_PMR_MASK) >> ICH_VMCR_PMR_SHIFT;
}
+#define INITIAL_PENDBASER_VALUE \
+ (GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWb) | \
+ GIC_BASER_CACHEABILITY(GICR_PENDBASER, OUTER, SameAsInner) | \
+ GIC_BASER_SHAREABILITY(GICR_PENDBASER, InnerShareable))
+
void vgic_v3_enable(struct kvm_vcpu *vcpu)
{
struct vgic_v3_cpu_if *vgic_v3 = &vcpu->arch.vgic_cpu.vgic_v3;
@@ -208,10 +213,12 @@ void vgic_v3_enable(struct kvm_vcpu *vcpu)
* way, so we force SRE to 1 to demonstrate this to the guest.
* This goes with the spec allowing the value to be RAO/WI.
*/
- if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3)
+ if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) {
vgic_v3->vgic_sre = ICC_SRE_EL1_SRE;
- else
+ vcpu->arch.vgic_cpu.pendbaser = INITIAL_PENDBASER_VALUE;
+ } else {
vgic_v3->vgic_sre = 0;
+ }
/* Get the show on the road... */
vgic_v3->vgic_hcr = ICH_HCR_EN;
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-07-22 17:28 ` [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers Marc Zyngier
@ 2016-08-01 18:20 ` Christoffer Dall
2016-08-02 9:40 ` Andre Przywara
0 siblings, 1 reply; 67+ messages in thread
From: Christoffer Dall @ 2016-08-01 18:20 UTC (permalink / raw)
To: linux-arm-kernel
On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
> From: Andre Przywara <andre.przywara@arm.com>
>
> In the GICv3 redistributor there are the PENDBASER and PROPBASER
> registers which we did not emulate so far, as they only make sense
> when having an ITS. In preparation for that emulate those MMIO
> accesses by storing the 64-bit data written into it into a variable
> which we later read in the ITS emulation.
> We also sanitise the registers, making sure RES0 regions are respected
> and checking for valid memory attributes.
>
> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
> Tested-by: Eric Auger <eric.auger@redhat.com>
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> ---
> include/kvm/arm_vgic.h | 13 ++++
> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
> 4 files changed, 181 insertions(+), 4 deletions(-)
>
> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
> index 450b4da..df2dec5 100644
> --- a/include/kvm/arm_vgic.h
> +++ b/include/kvm/arm_vgic.h
> @@ -146,6 +146,14 @@ struct vgic_dist {
> struct vgic_irq *spis;
>
> struct vgic_io_device dist_iodev;
> +
> + /*
> + * Contains the attributes and gpa of the LPI configuration table.
> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
> + * one address across all redistributors.
> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
> + */
> + u64 propbaser;
> };
>
> struct vgic_v2_cpu_if {
> @@ -200,6 +208,11 @@ struct vgic_cpu {
> */
> struct vgic_io_device rd_iodev;
> struct vgic_io_device sgi_iodev;
> +
> + /* Contains the attributes and gpa of the LPI pending tables. */
> + u64 pendbaser;
> +
> + bool lpis_enabled;
> };
>
> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
> index bfcafbd..278bfbb 100644
> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
> }
>
> +/* allows updates of any half of a 64-bit register (or the whole thing) */
> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
> + unsigned long val)
> +{
> + int lower = (offset & 4) * 8;
> + int upper = lower + 8 * len - 1;
> +
> + reg &= ~GENMASK_ULL(upper, lower);
> + val &= GENMASK_ULL(len * 8 - 1, 0);
> +
> + return reg | ((u64)val << lower);
> +}
> +
> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
> gpa_t addr, unsigned int len)
> {
> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
> return 0;
> }
>
> +/* We want to avoid outer shareable. */
> +u64 vgic_sanitise_shareability(u64 field)
> +{
> + switch (field) {
> + case GIC_BASER_OuterShareable:
> + return GIC_BASER_InnerShareable;
> + default:
> + return field;
> + }
> +}
> +
> +/* Avoid any inner non-cacheable mapping. */
> +u64 vgic_sanitise_inner_cacheability(u64 field)
> +{
> + switch (field) {
> + case GIC_BASER_CACHE_nCnB:
> + case GIC_BASER_CACHE_nC:
> + return GIC_BASER_CACHE_RaWb;
> + default:
> + return field;
> + }
> +}
> +
> +/* Non-cacheable or same-as-inner are OK. */
> +u64 vgic_sanitise_outer_cacheability(u64 field)
> +{
> + switch (field) {
> + case GIC_BASER_CACHE_SameAsInner:
> + case GIC_BASER_CACHE_nC:
> + return field;
> + default:
> + return GIC_BASER_CACHE_nC;
> + }
> +}
> +
> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
> + u64 (*sanitise_fn)(u64))
> +{
> + u64 field = (reg & field_mask) >> field_shift;
> +
> + field = sanitise_fn(field) << field_shift;
> + return (reg & ~field_mask) | field;
> +}
> +
> +#define PROPBASER_RES0_MASK \
> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
> +#define PENDBASER_RES0_MASK \
> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
> +
> +static u64 vgic_sanitise_pendbaser(u64 reg)
> +{
> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
> + GICR_PENDBASER_SHAREABILITY_SHIFT,
> + vgic_sanitise_shareability);
> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
> + vgic_sanitise_inner_cacheability);
> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
> + vgic_sanitise_outer_cacheability);
> +
> + reg &= ~PENDBASER_RES0_MASK;
> + reg &= ~GENMASK_ULL(51, 48);
> +
> + return reg;
> +}
> +
> +static u64 vgic_sanitise_propbaser(u64 reg)
> +{
> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
> + GICR_PROPBASER_SHAREABILITY_SHIFT,
> + vgic_sanitise_shareability);
> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
> + vgic_sanitise_inner_cacheability);
> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
> + vgic_sanitise_outer_cacheability);
> +
> + reg &= ~PROPBASER_RES0_MASK;
> + reg &= ~GENMASK_ULL(51, 48);
> + return reg;
> +}
> +
> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
> + gpa_t addr, unsigned int len)
> +{
> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
> +
> + return extract_bytes(dist->propbaser, addr & 7, len);
> +}
> +
> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
> + gpa_t addr, unsigned int len,
> + unsigned long val)
> +{
> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
> + u64 propbaser = dist->propbaser;
> +
> + /* Storing a value with LPIs already enabled is undefined */
> + if (vgic_cpu->lpis_enabled)
> + return;
> +
> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
> + propbaser = vgic_sanitise_propbaser(propbaser);
> +
> + dist->propbaser = propbaser;
Which guarantees do we have that this will always be a single atomic
write?
> +}
> +
> +static unsigned long vgic_mmio_read_pendbase(struct kvm_vcpu *vcpu,
> + gpa_t addr, unsigned int len)
> +{
> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
> +
> + return extract_bytes(vgic_cpu->pendbaser, addr & 7, len);
> +}
> +
> +static void vgic_mmio_write_pendbase(struct kvm_vcpu *vcpu,
> + gpa_t addr, unsigned int len,
> + unsigned long val)
> +{
> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
> + u64 pendbaser = vgic_cpu->pendbaser;
> +
> + /* Storing a value with LPIs already enabled is undefined */
> + if (vgic_cpu->lpis_enabled)
> + return;
> +
> + pendbaser = update_64bit_reg(pendbaser, addr & 4, len, val);
> + pendbaser = vgic_sanitise_pendbaser(pendbaser);
> +
> + vgic_cpu->pendbaser = pendbaser;
same (assuming multiple VCPUs could be accessing this redistributor at
the same time?)
> +}
> +
> /*
> * The GICv3 per-IRQ registers are split to control PPIs and SGIs in the
> * redistributors, while SPIs are covered by registers in the distributor
> @@ -232,10 +381,10 @@ static const struct vgic_register_region vgic_v3_rdbase_registers[] = {
> vgic_mmio_read_v3r_typer, vgic_mmio_write_wi, 8,
> VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
> REGISTER_DESC_WITH_LENGTH(GICR_PROPBASER,
> - vgic_mmio_read_raz, vgic_mmio_write_wi, 8,
> + vgic_mmio_read_propbase, vgic_mmio_write_propbase, 8,
> VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
> REGISTER_DESC_WITH_LENGTH(GICR_PENDBASER,
> - vgic_mmio_read_raz, vgic_mmio_write_wi, 8,
> + vgic_mmio_read_pendbase, vgic_mmio_write_pendbase, 8,
> VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
> REGISTER_DESC_WITH_LENGTH(GICR_IDREGS,
> vgic_mmio_read_v3_idregs, vgic_mmio_write_wi, 48,
> diff --git a/virt/kvm/arm/vgic/vgic-mmio.h b/virt/kvm/arm/vgic/vgic-mmio.h
> index 8509014..71aa39d 100644
> --- a/virt/kvm/arm/vgic/vgic-mmio.h
> +++ b/virt/kvm/arm/vgic/vgic-mmio.h
> @@ -147,4 +147,12 @@ unsigned int vgic_v2_init_dist_iodev(struct vgic_io_device *dev);
>
> unsigned int vgic_v3_init_dist_iodev(struct vgic_io_device *dev);
>
> +#ifdef CONFIG_KVM_ARM_VGIC_V3
> +u64 vgic_sanitise_outer_cacheability(u64 reg);
> +u64 vgic_sanitise_inner_cacheability(u64 reg);
> +u64 vgic_sanitise_shareability(u64 reg);
> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
> + u64 (*sanitise_fn)(u64));
> +#endif
> +
> #endif
> diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c
> index f0ac064..6f8f31f 100644
> --- a/virt/kvm/arm/vgic/vgic-v3.c
> +++ b/virt/kvm/arm/vgic/vgic-v3.c
> @@ -191,6 +191,11 @@ void vgic_v3_get_vmcr(struct kvm_vcpu *vcpu, struct vgic_vmcr *vmcrp)
> vmcrp->pmr = (vmcr & ICH_VMCR_PMR_MASK) >> ICH_VMCR_PMR_SHIFT;
> }
>
> +#define INITIAL_PENDBASER_VALUE \
> + (GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWb) | \
> + GIC_BASER_CACHEABILITY(GICR_PENDBASER, OUTER, SameAsInner) | \
> + GIC_BASER_SHAREABILITY(GICR_PENDBASER, InnerShareable))
> +
> void vgic_v3_enable(struct kvm_vcpu *vcpu)
> {
> struct vgic_v3_cpu_if *vgic_v3 = &vcpu->arch.vgic_cpu.vgic_v3;
> @@ -208,10 +213,12 @@ void vgic_v3_enable(struct kvm_vcpu *vcpu)
> * way, so we force SRE to 1 to demonstrate this to the guest.
> * This goes with the spec allowing the value to be RAO/WI.
> */
> - if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3)
> + if (vcpu->kvm->arch.vgic.vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) {
> vgic_v3->vgic_sre = ICC_SRE_EL1_SRE;
> - else
> + vcpu->arch.vgic_cpu.pendbaser = INITIAL_PENDBASER_VALUE;
> + } else {
> vgic_v3->vgic_sre = 0;
> + }
>
> /* Get the show on the road... */
> vgic_v3->vgic_hcr = ICH_HCR_EN;
> --
> 2.8.1
>
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-08-01 18:20 ` Christoffer Dall
@ 2016-08-02 9:40 ` Andre Przywara
2016-08-02 10:12 ` Marc Zyngier
0 siblings, 1 reply; 67+ messages in thread
From: Andre Przywara @ 2016-08-02 9:40 UTC (permalink / raw)
To: linux-arm-kernel
Hi,
On 01/08/16 19:20, Christoffer Dall wrote:
> On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
>> From: Andre Przywara <andre.przywara@arm.com>
>>
>> In the GICv3 redistributor there are the PENDBASER and PROPBASER
>> registers which we did not emulate so far, as they only make sense
>> when having an ITS. In preparation for that emulate those MMIO
>> accesses by storing the 64-bit data written into it into a variable
>> which we later read in the ITS emulation.
>> We also sanitise the registers, making sure RES0 regions are respected
>> and checking for valid memory attributes.
>>
>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
>> Tested-by: Eric Auger <eric.auger@redhat.com>
>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
>> ---
>> include/kvm/arm_vgic.h | 13 ++++
>> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
>> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
>> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
>> 4 files changed, 181 insertions(+), 4 deletions(-)
>>
>> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
>> index 450b4da..df2dec5 100644
>> --- a/include/kvm/arm_vgic.h
>> +++ b/include/kvm/arm_vgic.h
>> @@ -146,6 +146,14 @@ struct vgic_dist {
>> struct vgic_irq *spis;
>>
>> struct vgic_io_device dist_iodev;
>> +
>> + /*
>> + * Contains the attributes and gpa of the LPI configuration table.
>> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
>> + * one address across all redistributors.
>> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
>> + */
>> + u64 propbaser;
>> };
>>
>> struct vgic_v2_cpu_if {
>> @@ -200,6 +208,11 @@ struct vgic_cpu {
>> */
>> struct vgic_io_device rd_iodev;
>> struct vgic_io_device sgi_iodev;
>> +
>> + /* Contains the attributes and gpa of the LPI pending tables. */
>> + u64 pendbaser;
>> +
>> + bool lpis_enabled;
>> };
>>
>> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
>> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>> index bfcafbd..278bfbb 100644
>> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
>> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
>> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
>> }
>>
>> +/* allows updates of any half of a 64-bit register (or the whole thing) */
>> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
>> + unsigned long val)
>> +{
>> + int lower = (offset & 4) * 8;
>> + int upper = lower + 8 * len - 1;
>> +
>> + reg &= ~GENMASK_ULL(upper, lower);
>> + val &= GENMASK_ULL(len * 8 - 1, 0);
>> +
>> + return reg | ((u64)val << lower);
>> +}
>> +
>> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
>> gpa_t addr, unsigned int len)
>> {
>> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
>> return 0;
>> }
>>
>> +/* We want to avoid outer shareable. */
>> +u64 vgic_sanitise_shareability(u64 field)
>> +{
>> + switch (field) {
>> + case GIC_BASER_OuterShareable:
>> + return GIC_BASER_InnerShareable;
>> + default:
>> + return field;
>> + }
>> +}
>> +
>> +/* Avoid any inner non-cacheable mapping. */
>> +u64 vgic_sanitise_inner_cacheability(u64 field)
>> +{
>> + switch (field) {
>> + case GIC_BASER_CACHE_nCnB:
>> + case GIC_BASER_CACHE_nC:
>> + return GIC_BASER_CACHE_RaWb;
>> + default:
>> + return field;
>> + }
>> +}
>> +
>> +/* Non-cacheable or same-as-inner are OK. */
>> +u64 vgic_sanitise_outer_cacheability(u64 field)
>> +{
>> + switch (field) {
>> + case GIC_BASER_CACHE_SameAsInner:
>> + case GIC_BASER_CACHE_nC:
>> + return field;
>> + default:
>> + return GIC_BASER_CACHE_nC;
>> + }
>> +}
>> +
>> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
>> + u64 (*sanitise_fn)(u64))
>> +{
>> + u64 field = (reg & field_mask) >> field_shift;
>> +
>> + field = sanitise_fn(field) << field_shift;
>> + return (reg & ~field_mask) | field;
>> +}
>> +
>> +#define PROPBASER_RES0_MASK \
>> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
>> +#define PENDBASER_RES0_MASK \
>> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
>> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
>> +
>> +static u64 vgic_sanitise_pendbaser(u64 reg)
>> +{
>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
>> + GICR_PENDBASER_SHAREABILITY_SHIFT,
>> + vgic_sanitise_shareability);
>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
>> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
>> + vgic_sanitise_inner_cacheability);
>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
>> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
>> + vgic_sanitise_outer_cacheability);
>> +
>> + reg &= ~PENDBASER_RES0_MASK;
>> + reg &= ~GENMASK_ULL(51, 48);
>> +
>> + return reg;
>> +}
>> +
>> +static u64 vgic_sanitise_propbaser(u64 reg)
>> +{
>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
>> + GICR_PROPBASER_SHAREABILITY_SHIFT,
>> + vgic_sanitise_shareability);
>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
>> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
>> + vgic_sanitise_inner_cacheability);
>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
>> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
>> + vgic_sanitise_outer_cacheability);
>> +
>> + reg &= ~PROPBASER_RES0_MASK;
>> + reg &= ~GENMASK_ULL(51, 48);
>> + return reg;
>> +}
>> +
>> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
>> + gpa_t addr, unsigned int len)
>> +{
>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>> +
>> + return extract_bytes(dist->propbaser, addr & 7, len);
>> +}
>> +
>> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
>> + gpa_t addr, unsigned int len,
>> + unsigned long val)
>> +{
>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>> + u64 propbaser = dist->propbaser;
>> +
>> + /* Storing a value with LPIs already enabled is undefined */
>> + if (vgic_cpu->lpis_enabled)
>> + return;
>> +
>> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
>> + propbaser = vgic_sanitise_propbaser(propbaser);
>> +
>> + dist->propbaser = propbaser;
>
> Which guarantees do we have that this will always be a single atomic
> write?
At least on arm64 - which is the only architecture this code is
compiling for at the moment - I don't see why this shouldn't be a single
write:
str x19, [x22,#3544]
is what my setup here creates.
Do we need something stronger? Do we want to postpone this to the point
when we get arm(32) support?
>
>> +}
>> +
>> +static unsigned long vgic_mmio_read_pendbase(struct kvm_vcpu *vcpu,
>> + gpa_t addr, unsigned int len)
>> +{
>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>> +
>> + return extract_bytes(vgic_cpu->pendbaser, addr & 7, len);
>> +}
>> +
>> +static void vgic_mmio_write_pendbase(struct kvm_vcpu *vcpu,
>> + gpa_t addr, unsigned int len,
>> + unsigned long val)
>> +{
>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>> + u64 pendbaser = vgic_cpu->pendbaser;
>> +
>> + /* Storing a value with LPIs already enabled is undefined */
>> + if (vgic_cpu->lpis_enabled)
>> + return;
>> +
>> + pendbaser = update_64bit_reg(pendbaser, addr & 4, len, val);
>> + pendbaser = vgic_sanitise_pendbaser(pendbaser);
>> +
>> + vgic_cpu->pendbaser = pendbaser;
>
> same (assuming multiple VCPUs could be accessing this redistributor at
> the same time?)
same answer ;-)
Cheers,
Andre.
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-08-02 9:40 ` Andre Przywara
@ 2016-08-02 10:12 ` Marc Zyngier
2016-08-02 14:33 ` Christoffer Dall
0 siblings, 1 reply; 67+ messages in thread
From: Marc Zyngier @ 2016-08-02 10:12 UTC (permalink / raw)
To: linux-arm-kernel
On 02/08/16 10:40, Andre Przywara wrote:
> Hi,
>
> On 01/08/16 19:20, Christoffer Dall wrote:
>> On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
>>> From: Andre Przywara <andre.przywara@arm.com>
>>>
>>> In the GICv3 redistributor there are the PENDBASER and PROPBASER
>>> registers which we did not emulate so far, as they only make sense
>>> when having an ITS. In preparation for that emulate those MMIO
>>> accesses by storing the 64-bit data written into it into a variable
>>> which we later read in the ITS emulation.
>>> We also sanitise the registers, making sure RES0 regions are respected
>>> and checking for valid memory attributes.
>>>
>>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
>>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
>>> Tested-by: Eric Auger <eric.auger@redhat.com>
>>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
>>> ---
>>> include/kvm/arm_vgic.h | 13 ++++
>>> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
>>> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
>>> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
>>> 4 files changed, 181 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
>>> index 450b4da..df2dec5 100644
>>> --- a/include/kvm/arm_vgic.h
>>> +++ b/include/kvm/arm_vgic.h
>>> @@ -146,6 +146,14 @@ struct vgic_dist {
>>> struct vgic_irq *spis;
>>>
>>> struct vgic_io_device dist_iodev;
>>> +
>>> + /*
>>> + * Contains the attributes and gpa of the LPI configuration table.
>>> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
>>> + * one address across all redistributors.
>>> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
>>> + */
>>> + u64 propbaser;
>>> };
>>>
>>> struct vgic_v2_cpu_if {
>>> @@ -200,6 +208,11 @@ struct vgic_cpu {
>>> */
>>> struct vgic_io_device rd_iodev;
>>> struct vgic_io_device sgi_iodev;
>>> +
>>> + /* Contains the attributes and gpa of the LPI pending tables. */
>>> + u64 pendbaser;
>>> +
>>> + bool lpis_enabled;
>>> };
>>>
>>> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
>>> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>> index bfcafbd..278bfbb 100644
>>> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
>>> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
>>> }
>>>
>>> +/* allows updates of any half of a 64-bit register (or the whole thing) */
>>> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
>>> + unsigned long val)
>>> +{
>>> + int lower = (offset & 4) * 8;
>>> + int upper = lower + 8 * len - 1;
>>> +
>>> + reg &= ~GENMASK_ULL(upper, lower);
>>> + val &= GENMASK_ULL(len * 8 - 1, 0);
>>> +
>>> + return reg | ((u64)val << lower);
>>> +}
>>> +
>>> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
>>> gpa_t addr, unsigned int len)
>>> {
>>> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
>>> return 0;
>>> }
>>>
>>> +/* We want to avoid outer shareable. */
>>> +u64 vgic_sanitise_shareability(u64 field)
>>> +{
>>> + switch (field) {
>>> + case GIC_BASER_OuterShareable:
>>> + return GIC_BASER_InnerShareable;
>>> + default:
>>> + return field;
>>> + }
>>> +}
>>> +
>>> +/* Avoid any inner non-cacheable mapping. */
>>> +u64 vgic_sanitise_inner_cacheability(u64 field)
>>> +{
>>> + switch (field) {
>>> + case GIC_BASER_CACHE_nCnB:
>>> + case GIC_BASER_CACHE_nC:
>>> + return GIC_BASER_CACHE_RaWb;
>>> + default:
>>> + return field;
>>> + }
>>> +}
>>> +
>>> +/* Non-cacheable or same-as-inner are OK. */
>>> +u64 vgic_sanitise_outer_cacheability(u64 field)
>>> +{
>>> + switch (field) {
>>> + case GIC_BASER_CACHE_SameAsInner:
>>> + case GIC_BASER_CACHE_nC:
>>> + return field;
>>> + default:
>>> + return GIC_BASER_CACHE_nC;
>>> + }
>>> +}
>>> +
>>> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
>>> + u64 (*sanitise_fn)(u64))
>>> +{
>>> + u64 field = (reg & field_mask) >> field_shift;
>>> +
>>> + field = sanitise_fn(field) << field_shift;
>>> + return (reg & ~field_mask) | field;
>>> +}
>>> +
>>> +#define PROPBASER_RES0_MASK \
>>> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
>>> +#define PENDBASER_RES0_MASK \
>>> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
>>> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
>>> +
>>> +static u64 vgic_sanitise_pendbaser(u64 reg)
>>> +{
>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
>>> + GICR_PENDBASER_SHAREABILITY_SHIFT,
>>> + vgic_sanitise_shareability);
>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
>>> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
>>> + vgic_sanitise_inner_cacheability);
>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
>>> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
>>> + vgic_sanitise_outer_cacheability);
>>> +
>>> + reg &= ~PENDBASER_RES0_MASK;
>>> + reg &= ~GENMASK_ULL(51, 48);
>>> +
>>> + return reg;
>>> +}
>>> +
>>> +static u64 vgic_sanitise_propbaser(u64 reg)
>>> +{
>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
>>> + GICR_PROPBASER_SHAREABILITY_SHIFT,
>>> + vgic_sanitise_shareability);
>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
>>> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
>>> + vgic_sanitise_inner_cacheability);
>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
>>> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
>>> + vgic_sanitise_outer_cacheability);
>>> +
>>> + reg &= ~PROPBASER_RES0_MASK;
>>> + reg &= ~GENMASK_ULL(51, 48);
>>> + return reg;
>>> +}
>>> +
>>> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
>>> + gpa_t addr, unsigned int len)
>>> +{
>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>>> +
>>> + return extract_bytes(dist->propbaser, addr & 7, len);
>>> +}
>>> +
>>> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
>>> + gpa_t addr, unsigned int len,
>>> + unsigned long val)
>>> +{
>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>>> + u64 propbaser = dist->propbaser;
>>> +
>>> + /* Storing a value with LPIs already enabled is undefined */
>>> + if (vgic_cpu->lpis_enabled)
>>> + return;
>>> +
>>> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
>>> + propbaser = vgic_sanitise_propbaser(propbaser);
>>> +
>>> + dist->propbaser = propbaser;
>>
>> Which guarantees do we have that this will always be a single atomic
>> write?
>
> At least on arm64 - which is the only architecture this code is
> compiling for at the moment - I don't see why this shouldn't be a single
> write:
> str x19, [x22,#3544]
> is what my setup here creates.
>
> Do we need something stronger? Do we want to postpone this to the point
> when we get arm(32) support?
This is a *device*. It shouldn't be affected by whatever drives it.
Now, and more to the point: the write shouldn't have to be atomic. All
64bit registers should be able to cope with 32bit writes to it, as
described in the architecture spec (IHI0069C 8.1.3).
The important thing to ensure is that we don't use that value as long it
can change, which means that we can't use it as long as LPIs are
disabled. Which means that things like update_lpi_config() shouldn't
even try and read from memory if LPIs are not enabled.
>
>>
>>> +}
>>> +
>>> +static unsigned long vgic_mmio_read_pendbase(struct kvm_vcpu *vcpu,
>>> + gpa_t addr, unsigned int len)
>>> +{
>>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>>> +
>>> + return extract_bytes(vgic_cpu->pendbaser, addr & 7, len);
>>> +}
>>> +
>>> +static void vgic_mmio_write_pendbase(struct kvm_vcpu *vcpu,
>>> + gpa_t addr, unsigned int len,
>>> + unsigned long val)
>>> +{
>>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>>> + u64 pendbaser = vgic_cpu->pendbaser;
>>> +
>>> + /* Storing a value with LPIs already enabled is undefined */
>>> + if (vgic_cpu->lpis_enabled)
>>> + return;
>>> +
>>> + pendbaser = update_64bit_reg(pendbaser, addr & 4, len, val);
>>> + pendbaser = vgic_sanitise_pendbaser(pendbaser);
>>> +
>>> + vgic_cpu->pendbaser = pendbaser;
>>
>> same (assuming multiple VCPUs could be accessing this redistributor at
>> the same time?)
>
> same answer ;-)
Same as well. its_sync_lpi_pending_table() shouldn't sync anything until
lpis_enabled is set.
Thanks,
M.
--
Jazz is not dead. It just smells funny...
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-08-02 10:12 ` Marc Zyngier
@ 2016-08-02 14:33 ` Christoffer Dall
2016-08-02 14:46 ` Marc Zyngier
0 siblings, 1 reply; 67+ messages in thread
From: Christoffer Dall @ 2016-08-02 14:33 UTC (permalink / raw)
To: linux-arm-kernel
On Tue, Aug 02, 2016 at 11:12:46AM +0100, Marc Zyngier wrote:
> On 02/08/16 10:40, Andre Przywara wrote:
> > Hi,
> >
> > On 01/08/16 19:20, Christoffer Dall wrote:
> >> On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
> >>> From: Andre Przywara <andre.przywara@arm.com>
> >>>
> >>> In the GICv3 redistributor there are the PENDBASER and PROPBASER
> >>> registers which we did not emulate so far, as they only make sense
> >>> when having an ITS. In preparation for that emulate those MMIO
> >>> accesses by storing the 64-bit data written into it into a variable
> >>> which we later read in the ITS emulation.
> >>> We also sanitise the registers, making sure RES0 regions are respected
> >>> and checking for valid memory attributes.
> >>>
> >>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
> >>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
> >>> Tested-by: Eric Auger <eric.auger@redhat.com>
> >>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> >>> ---
> >>> include/kvm/arm_vgic.h | 13 ++++
> >>> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
> >>> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
> >>> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
> >>> 4 files changed, 181 insertions(+), 4 deletions(-)
> >>>
> >>> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
> >>> index 450b4da..df2dec5 100644
> >>> --- a/include/kvm/arm_vgic.h
> >>> +++ b/include/kvm/arm_vgic.h
> >>> @@ -146,6 +146,14 @@ struct vgic_dist {
> >>> struct vgic_irq *spis;
> >>>
> >>> struct vgic_io_device dist_iodev;
> >>> +
> >>> + /*
> >>> + * Contains the attributes and gpa of the LPI configuration table.
> >>> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
> >>> + * one address across all redistributors.
> >>> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
> >>> + */
> >>> + u64 propbaser;
> >>> };
> >>>
> >>> struct vgic_v2_cpu_if {
> >>> @@ -200,6 +208,11 @@ struct vgic_cpu {
> >>> */
> >>> struct vgic_io_device rd_iodev;
> >>> struct vgic_io_device sgi_iodev;
> >>> +
> >>> + /* Contains the attributes and gpa of the LPI pending tables. */
> >>> + u64 pendbaser;
> >>> +
> >>> + bool lpis_enabled;
> >>> };
> >>>
> >>> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
> >>> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
> >>> index bfcafbd..278bfbb 100644
> >>> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
> >>> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
> >>> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
> >>> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
> >>> }
> >>>
> >>> +/* allows updates of any half of a 64-bit register (or the whole thing) */
> >>> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
> >>> + unsigned long val)
> >>> +{
> >>> + int lower = (offset & 4) * 8;
> >>> + int upper = lower + 8 * len - 1;
> >>> +
> >>> + reg &= ~GENMASK_ULL(upper, lower);
> >>> + val &= GENMASK_ULL(len * 8 - 1, 0);
> >>> +
> >>> + return reg | ((u64)val << lower);
> >>> +}
> >>> +
> >>> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
> >>> gpa_t addr, unsigned int len)
> >>> {
> >>> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
> >>> return 0;
> >>> }
> >>>
> >>> +/* We want to avoid outer shareable. */
> >>> +u64 vgic_sanitise_shareability(u64 field)
> >>> +{
> >>> + switch (field) {
> >>> + case GIC_BASER_OuterShareable:
> >>> + return GIC_BASER_InnerShareable;
> >>> + default:
> >>> + return field;
> >>> + }
> >>> +}
> >>> +
> >>> +/* Avoid any inner non-cacheable mapping. */
> >>> +u64 vgic_sanitise_inner_cacheability(u64 field)
> >>> +{
> >>> + switch (field) {
> >>> + case GIC_BASER_CACHE_nCnB:
> >>> + case GIC_BASER_CACHE_nC:
> >>> + return GIC_BASER_CACHE_RaWb;
> >>> + default:
> >>> + return field;
> >>> + }
> >>> +}
> >>> +
> >>> +/* Non-cacheable or same-as-inner are OK. */
> >>> +u64 vgic_sanitise_outer_cacheability(u64 field)
> >>> +{
> >>> + switch (field) {
> >>> + case GIC_BASER_CACHE_SameAsInner:
> >>> + case GIC_BASER_CACHE_nC:
> >>> + return field;
> >>> + default:
> >>> + return GIC_BASER_CACHE_nC;
> >>> + }
> >>> +}
> >>> +
> >>> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
> >>> + u64 (*sanitise_fn)(u64))
> >>> +{
> >>> + u64 field = (reg & field_mask) >> field_shift;
> >>> +
> >>> + field = sanitise_fn(field) << field_shift;
> >>> + return (reg & ~field_mask) | field;
> >>> +}
> >>> +
> >>> +#define PROPBASER_RES0_MASK \
> >>> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
> >>> +#define PENDBASER_RES0_MASK \
> >>> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
> >>> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
> >>> +
> >>> +static u64 vgic_sanitise_pendbaser(u64 reg)
> >>> +{
> >>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
> >>> + GICR_PENDBASER_SHAREABILITY_SHIFT,
> >>> + vgic_sanitise_shareability);
> >>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
> >>> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
> >>> + vgic_sanitise_inner_cacheability);
> >>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
> >>> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
> >>> + vgic_sanitise_outer_cacheability);
> >>> +
> >>> + reg &= ~PENDBASER_RES0_MASK;
> >>> + reg &= ~GENMASK_ULL(51, 48);
> >>> +
> >>> + return reg;
> >>> +}
> >>> +
> >>> +static u64 vgic_sanitise_propbaser(u64 reg)
> >>> +{
> >>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
> >>> + GICR_PROPBASER_SHAREABILITY_SHIFT,
> >>> + vgic_sanitise_shareability);
> >>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
> >>> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
> >>> + vgic_sanitise_inner_cacheability);
> >>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
> >>> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
> >>> + vgic_sanitise_outer_cacheability);
> >>> +
> >>> + reg &= ~PROPBASER_RES0_MASK;
> >>> + reg &= ~GENMASK_ULL(51, 48);
> >>> + return reg;
> >>> +}
> >>> +
> >>> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
> >>> + gpa_t addr, unsigned int len)
> >>> +{
> >>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
> >>> +
> >>> + return extract_bytes(dist->propbaser, addr & 7, len);
> >>> +}
> >>> +
> >>> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
> >>> + gpa_t addr, unsigned int len,
> >>> + unsigned long val)
> >>> +{
> >>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
> >>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
> >>> + u64 propbaser = dist->propbaser;
> >>> +
> >>> + /* Storing a value with LPIs already enabled is undefined */
> >>> + if (vgic_cpu->lpis_enabled)
> >>> + return;
> >>> +
> >>> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
> >>> + propbaser = vgic_sanitise_propbaser(propbaser);
> >>> +
> >>> + dist->propbaser = propbaser;
> >>
> >> Which guarantees do we have that this will always be a single atomic
> >> write?
> >
> > At least on arm64 - which is the only architecture this code is
> > compiling for at the moment - I don't see why this shouldn't be a single
> > write:
> > str x19, [x22,#3544]
> > is what my setup here creates.
> >
> > Do we need something stronger? Do we want to postpone this to the point
> > when we get arm(32) support?
>
> This is a *device*. It shouldn't be affected by whatever drives it.
>
> Now, and more to the point: the write shouldn't have to be atomic. All
> 64bit registers should be able to cope with 32bit writes to it, as
> described in the architecture spec (IHI0069C 8.1.3).
That's not my concern. My concern is that you have two CPUs updating
the propbaser, once after the other, and you end up with a mix of the
two updates. If this C-code can ever become two 32-bit writes, for
example, and these can happen in parallel you can end up with something
like that.
If there is not valid expectation from guest software that a real device
stores either one or the other value, then it's fine to leave it as is.
Thinking about it, any sane guest would probably synchronize multiple
writes to this register itself?
>
> The important thing to ensure is that we don't use that value as long it
> can change, which means that we can't use it as long as LPIs are
> disabled. Which means that things like update_lpi_config() shouldn't
> even try and read from memory if LPIs are not enabled.
>
I agree with this too, but that wasn't what my comment was targeting.
Thanks,
-Christoffer
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-08-02 14:33 ` Christoffer Dall
@ 2016-08-02 14:46 ` Marc Zyngier
2016-08-02 14:55 ` Christoffer Dall
0 siblings, 1 reply; 67+ messages in thread
From: Marc Zyngier @ 2016-08-02 14:46 UTC (permalink / raw)
To: linux-arm-kernel
On 02/08/16 15:33, Christoffer Dall wrote:
> On Tue, Aug 02, 2016 at 11:12:46AM +0100, Marc Zyngier wrote:
>> On 02/08/16 10:40, Andre Przywara wrote:
>>> Hi,
>>>
>>> On 01/08/16 19:20, Christoffer Dall wrote:
>>>> On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
>>>>> From: Andre Przywara <andre.przywara@arm.com>
>>>>>
>>>>> In the GICv3 redistributor there are the PENDBASER and PROPBASER
>>>>> registers which we did not emulate so far, as they only make sense
>>>>> when having an ITS. In preparation for that emulate those MMIO
>>>>> accesses by storing the 64-bit data written into it into a variable
>>>>> which we later read in the ITS emulation.
>>>>> We also sanitise the registers, making sure RES0 regions are respected
>>>>> and checking for valid memory attributes.
>>>>>
>>>>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
>>>>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
>>>>> Tested-by: Eric Auger <eric.auger@redhat.com>
>>>>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
>>>>> ---
>>>>> include/kvm/arm_vgic.h | 13 ++++
>>>>> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
>>>>> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
>>>>> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
>>>>> 4 files changed, 181 insertions(+), 4 deletions(-)
>>>>>
>>>>> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
>>>>> index 450b4da..df2dec5 100644
>>>>> --- a/include/kvm/arm_vgic.h
>>>>> +++ b/include/kvm/arm_vgic.h
>>>>> @@ -146,6 +146,14 @@ struct vgic_dist {
>>>>> struct vgic_irq *spis;
>>>>>
>>>>> struct vgic_io_device dist_iodev;
>>>>> +
>>>>> + /*
>>>>> + * Contains the attributes and gpa of the LPI configuration table.
>>>>> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
>>>>> + * one address across all redistributors.
>>>>> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
>>>>> + */
>>>>> + u64 propbaser;
>>>>> };
>>>>>
>>>>> struct vgic_v2_cpu_if {
>>>>> @@ -200,6 +208,11 @@ struct vgic_cpu {
>>>>> */
>>>>> struct vgic_io_device rd_iodev;
>>>>> struct vgic_io_device sgi_iodev;
>>>>> +
>>>>> + /* Contains the attributes and gpa of the LPI pending tables. */
>>>>> + u64 pendbaser;
>>>>> +
>>>>> + bool lpis_enabled;
>>>>> };
>>>>>
>>>>> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
>>>>> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>>>> index bfcafbd..278bfbb 100644
>>>>> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>>>> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>>>> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
>>>>> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
>>>>> }
>>>>>
>>>>> +/* allows updates of any half of a 64-bit register (or the whole thing) */
>>>>> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
>>>>> + unsigned long val)
>>>>> +{
>>>>> + int lower = (offset & 4) * 8;
>>>>> + int upper = lower + 8 * len - 1;
>>>>> +
>>>>> + reg &= ~GENMASK_ULL(upper, lower);
>>>>> + val &= GENMASK_ULL(len * 8 - 1, 0);
>>>>> +
>>>>> + return reg | ((u64)val << lower);
>>>>> +}
>>>>> +
>>>>> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
>>>>> gpa_t addr, unsigned int len)
>>>>> {
>>>>> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
>>>>> return 0;
>>>>> }
>>>>>
>>>>> +/* We want to avoid outer shareable. */
>>>>> +u64 vgic_sanitise_shareability(u64 field)
>>>>> +{
>>>>> + switch (field) {
>>>>> + case GIC_BASER_OuterShareable:
>>>>> + return GIC_BASER_InnerShareable;
>>>>> + default:
>>>>> + return field;
>>>>> + }
>>>>> +}
>>>>> +
>>>>> +/* Avoid any inner non-cacheable mapping. */
>>>>> +u64 vgic_sanitise_inner_cacheability(u64 field)
>>>>> +{
>>>>> + switch (field) {
>>>>> + case GIC_BASER_CACHE_nCnB:
>>>>> + case GIC_BASER_CACHE_nC:
>>>>> + return GIC_BASER_CACHE_RaWb;
>>>>> + default:
>>>>> + return field;
>>>>> + }
>>>>> +}
>>>>> +
>>>>> +/* Non-cacheable or same-as-inner are OK. */
>>>>> +u64 vgic_sanitise_outer_cacheability(u64 field)
>>>>> +{
>>>>> + switch (field) {
>>>>> + case GIC_BASER_CACHE_SameAsInner:
>>>>> + case GIC_BASER_CACHE_nC:
>>>>> + return field;
>>>>> + default:
>>>>> + return GIC_BASER_CACHE_nC;
>>>>> + }
>>>>> +}
>>>>> +
>>>>> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
>>>>> + u64 (*sanitise_fn)(u64))
>>>>> +{
>>>>> + u64 field = (reg & field_mask) >> field_shift;
>>>>> +
>>>>> + field = sanitise_fn(field) << field_shift;
>>>>> + return (reg & ~field_mask) | field;
>>>>> +}
>>>>> +
>>>>> +#define PROPBASER_RES0_MASK \
>>>>> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
>>>>> +#define PENDBASER_RES0_MASK \
>>>>> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
>>>>> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
>>>>> +
>>>>> +static u64 vgic_sanitise_pendbaser(u64 reg)
>>>>> +{
>>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
>>>>> + GICR_PENDBASER_SHAREABILITY_SHIFT,
>>>>> + vgic_sanitise_shareability);
>>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
>>>>> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
>>>>> + vgic_sanitise_inner_cacheability);
>>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
>>>>> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
>>>>> + vgic_sanitise_outer_cacheability);
>>>>> +
>>>>> + reg &= ~PENDBASER_RES0_MASK;
>>>>> + reg &= ~GENMASK_ULL(51, 48);
>>>>> +
>>>>> + return reg;
>>>>> +}
>>>>> +
>>>>> +static u64 vgic_sanitise_propbaser(u64 reg)
>>>>> +{
>>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
>>>>> + GICR_PROPBASER_SHAREABILITY_SHIFT,
>>>>> + vgic_sanitise_shareability);
>>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
>>>>> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
>>>>> + vgic_sanitise_inner_cacheability);
>>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
>>>>> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
>>>>> + vgic_sanitise_outer_cacheability);
>>>>> +
>>>>> + reg &= ~PROPBASER_RES0_MASK;
>>>>> + reg &= ~GENMASK_ULL(51, 48);
>>>>> + return reg;
>>>>> +}
>>>>> +
>>>>> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
>>>>> + gpa_t addr, unsigned int len)
>>>>> +{
>>>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>>>>> +
>>>>> + return extract_bytes(dist->propbaser, addr & 7, len);
>>>>> +}
>>>>> +
>>>>> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
>>>>> + gpa_t addr, unsigned int len,
>>>>> + unsigned long val)
>>>>> +{
>>>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>>>>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>>>>> + u64 propbaser = dist->propbaser;
>>>>> +
>>>>> + /* Storing a value with LPIs already enabled is undefined */
>>>>> + if (vgic_cpu->lpis_enabled)
>>>>> + return;
>>>>> +
>>>>> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
>>>>> + propbaser = vgic_sanitise_propbaser(propbaser);
>>>>> +
>>>>> + dist->propbaser = propbaser;
>>>>
>>>> Which guarantees do we have that this will always be a single atomic
>>>> write?
>>>
>>> At least on arm64 - which is the only architecture this code is
>>> compiling for at the moment - I don't see why this shouldn't be a single
>>> write:
>>> str x19, [x22,#3544]
>>> is what my setup here creates.
>>>
>>> Do we need something stronger? Do we want to postpone this to the point
>>> when we get arm(32) support?
>>
>> This is a *device*. It shouldn't be affected by whatever drives it.
>>
>> Now, and more to the point: the write shouldn't have to be atomic. All
>> 64bit registers should be able to cope with 32bit writes to it, as
>> described in the architecture spec (IHI0069C 8.1.3).
>
> That's not my concern. My concern is that you have two CPUs updating
> the propbaser, once after the other, and you end up with a mix of the
> two updates. If this C-code can ever become two 32-bit writes, for
> example, and these can happen in parallel you can end up with something
> like that.
>
> If there is not valid expectation from guest software that a real device
> stores either one or the other value, then it's fine to leave it as is.
>
> Thinking about it, any sane guest would probably synchronize multiple
> writes to this register itself?
Ah, I finally clocked what you mean. Yes, we need to guarantee that the
update to the register itself is atomic (architectural requirement).
Which means we need to turn this in to an atomic access (atomic64_set).
It will still be a normal store on arm64, and be a strd on arm, thanks
to having LPAE on the host.
Thanks,
M.
--
Jazz is not dead. It just smells funny...
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-08-02 14:46 ` Marc Zyngier
@ 2016-08-02 14:55 ` Christoffer Dall
2016-08-02 15:01 ` Marc Zyngier
0 siblings, 1 reply; 67+ messages in thread
From: Christoffer Dall @ 2016-08-02 14:55 UTC (permalink / raw)
To: linux-arm-kernel
On Tue, Aug 02, 2016 at 03:46:49PM +0100, Marc Zyngier wrote:
> On 02/08/16 15:33, Christoffer Dall wrote:
> > On Tue, Aug 02, 2016 at 11:12:46AM +0100, Marc Zyngier wrote:
> >> On 02/08/16 10:40, Andre Przywara wrote:
> >>> Hi,
> >>>
> >>> On 01/08/16 19:20, Christoffer Dall wrote:
> >>>> On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
> >>>>> From: Andre Przywara <andre.przywara@arm.com>
> >>>>>
> >>>>> In the GICv3 redistributor there are the PENDBASER and PROPBASER
> >>>>> registers which we did not emulate so far, as they only make sense
> >>>>> when having an ITS. In preparation for that emulate those MMIO
> >>>>> accesses by storing the 64-bit data written into it into a variable
> >>>>> which we later read in the ITS emulation.
> >>>>> We also sanitise the registers, making sure RES0 regions are respected
> >>>>> and checking for valid memory attributes.
> >>>>>
> >>>>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
> >>>>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
> >>>>> Tested-by: Eric Auger <eric.auger@redhat.com>
> >>>>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> >>>>> ---
> >>>>> include/kvm/arm_vgic.h | 13 ++++
> >>>>> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
> >>>>> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
> >>>>> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
> >>>>> 4 files changed, 181 insertions(+), 4 deletions(-)
> >>>>>
> >>>>> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
> >>>>> index 450b4da..df2dec5 100644
> >>>>> --- a/include/kvm/arm_vgic.h
> >>>>> +++ b/include/kvm/arm_vgic.h
> >>>>> @@ -146,6 +146,14 @@ struct vgic_dist {
> >>>>> struct vgic_irq *spis;
> >>>>>
> >>>>> struct vgic_io_device dist_iodev;
> >>>>> +
> >>>>> + /*
> >>>>> + * Contains the attributes and gpa of the LPI configuration table.
> >>>>> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
> >>>>> + * one address across all redistributors.
> >>>>> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
> >>>>> + */
> >>>>> + u64 propbaser;
> >>>>> };
> >>>>>
> >>>>> struct vgic_v2_cpu_if {
> >>>>> @@ -200,6 +208,11 @@ struct vgic_cpu {
> >>>>> */
> >>>>> struct vgic_io_device rd_iodev;
> >>>>> struct vgic_io_device sgi_iodev;
> >>>>> +
> >>>>> + /* Contains the attributes and gpa of the LPI pending tables. */
> >>>>> + u64 pendbaser;
> >>>>> +
> >>>>> + bool lpis_enabled;
> >>>>> };
> >>>>>
> >>>>> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
> >>>>> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
> >>>>> index bfcafbd..278bfbb 100644
> >>>>> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
> >>>>> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
> >>>>> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
> >>>>> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
> >>>>> }
> >>>>>
> >>>>> +/* allows updates of any half of a 64-bit register (or the whole thing) */
> >>>>> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
> >>>>> + unsigned long val)
> >>>>> +{
> >>>>> + int lower = (offset & 4) * 8;
> >>>>> + int upper = lower + 8 * len - 1;
> >>>>> +
> >>>>> + reg &= ~GENMASK_ULL(upper, lower);
> >>>>> + val &= GENMASK_ULL(len * 8 - 1, 0);
> >>>>> +
> >>>>> + return reg | ((u64)val << lower);
> >>>>> +}
> >>>>> +
> >>>>> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
> >>>>> gpa_t addr, unsigned int len)
> >>>>> {
> >>>>> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
> >>>>> return 0;
> >>>>> }
> >>>>>
> >>>>> +/* We want to avoid outer shareable. */
> >>>>> +u64 vgic_sanitise_shareability(u64 field)
> >>>>> +{
> >>>>> + switch (field) {
> >>>>> + case GIC_BASER_OuterShareable:
> >>>>> + return GIC_BASER_InnerShareable;
> >>>>> + default:
> >>>>> + return field;
> >>>>> + }
> >>>>> +}
> >>>>> +
> >>>>> +/* Avoid any inner non-cacheable mapping. */
> >>>>> +u64 vgic_sanitise_inner_cacheability(u64 field)
> >>>>> +{
> >>>>> + switch (field) {
> >>>>> + case GIC_BASER_CACHE_nCnB:
> >>>>> + case GIC_BASER_CACHE_nC:
> >>>>> + return GIC_BASER_CACHE_RaWb;
> >>>>> + default:
> >>>>> + return field;
> >>>>> + }
> >>>>> +}
> >>>>> +
> >>>>> +/* Non-cacheable or same-as-inner are OK. */
> >>>>> +u64 vgic_sanitise_outer_cacheability(u64 field)
> >>>>> +{
> >>>>> + switch (field) {
> >>>>> + case GIC_BASER_CACHE_SameAsInner:
> >>>>> + case GIC_BASER_CACHE_nC:
> >>>>> + return field;
> >>>>> + default:
> >>>>> + return GIC_BASER_CACHE_nC;
> >>>>> + }
> >>>>> +}
> >>>>> +
> >>>>> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
> >>>>> + u64 (*sanitise_fn)(u64))
> >>>>> +{
> >>>>> + u64 field = (reg & field_mask) >> field_shift;
> >>>>> +
> >>>>> + field = sanitise_fn(field) << field_shift;
> >>>>> + return (reg & ~field_mask) | field;
> >>>>> +}
> >>>>> +
> >>>>> +#define PROPBASER_RES0_MASK \
> >>>>> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
> >>>>> +#define PENDBASER_RES0_MASK \
> >>>>> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
> >>>>> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
> >>>>> +
> >>>>> +static u64 vgic_sanitise_pendbaser(u64 reg)
> >>>>> +{
> >>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
> >>>>> + GICR_PENDBASER_SHAREABILITY_SHIFT,
> >>>>> + vgic_sanitise_shareability);
> >>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
> >>>>> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
> >>>>> + vgic_sanitise_inner_cacheability);
> >>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
> >>>>> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
> >>>>> + vgic_sanitise_outer_cacheability);
> >>>>> +
> >>>>> + reg &= ~PENDBASER_RES0_MASK;
> >>>>> + reg &= ~GENMASK_ULL(51, 48);
> >>>>> +
> >>>>> + return reg;
> >>>>> +}
> >>>>> +
> >>>>> +static u64 vgic_sanitise_propbaser(u64 reg)
> >>>>> +{
> >>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
> >>>>> + GICR_PROPBASER_SHAREABILITY_SHIFT,
> >>>>> + vgic_sanitise_shareability);
> >>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
> >>>>> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
> >>>>> + vgic_sanitise_inner_cacheability);
> >>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
> >>>>> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
> >>>>> + vgic_sanitise_outer_cacheability);
> >>>>> +
> >>>>> + reg &= ~PROPBASER_RES0_MASK;
> >>>>> + reg &= ~GENMASK_ULL(51, 48);
> >>>>> + return reg;
> >>>>> +}
> >>>>> +
> >>>>> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
> >>>>> + gpa_t addr, unsigned int len)
> >>>>> +{
> >>>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
> >>>>> +
> >>>>> + return extract_bytes(dist->propbaser, addr & 7, len);
> >>>>> +}
> >>>>> +
> >>>>> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
> >>>>> + gpa_t addr, unsigned int len,
> >>>>> + unsigned long val)
> >>>>> +{
> >>>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
> >>>>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
> >>>>> + u64 propbaser = dist->propbaser;
> >>>>> +
> >>>>> + /* Storing a value with LPIs already enabled is undefined */
> >>>>> + if (vgic_cpu->lpis_enabled)
> >>>>> + return;
> >>>>> +
> >>>>> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
> >>>>> + propbaser = vgic_sanitise_propbaser(propbaser);
> >>>>> +
> >>>>> + dist->propbaser = propbaser;
> >>>>
> >>>> Which guarantees do we have that this will always be a single atomic
> >>>> write?
> >>>
> >>> At least on arm64 - which is the only architecture this code is
> >>> compiling for at the moment - I don't see why this shouldn't be a single
> >>> write:
> >>> str x19, [x22,#3544]
> >>> is what my setup here creates.
> >>>
> >>> Do we need something stronger? Do we want to postpone this to the point
> >>> when we get arm(32) support?
> >>
> >> This is a *device*. It shouldn't be affected by whatever drives it.
> >>
> >> Now, and more to the point: the write shouldn't have to be atomic. All
> >> 64bit registers should be able to cope with 32bit writes to it, as
> >> described in the architecture spec (IHI0069C 8.1.3).
> >
> > That's not my concern. My concern is that you have two CPUs updating
> > the propbaser, once after the other, and you end up with a mix of the
> > two updates. If this C-code can ever become two 32-bit writes, for
> > example, and these can happen in parallel you can end up with something
> > like that.
> >
> > If there is not valid expectation from guest software that a real device
> > stores either one or the other value, then it's fine to leave it as is.
> >
> > Thinking about it, any sane guest would probably synchronize multiple
> > writes to this register itself?
>
> Ah, I finally clocked what you mean. Yes, we need to guarantee that the
> update to the register itself is atomic (architectural requirement).
> Which means we need to turn this in to an atomic access (atomic64_set).
Doesn't atomic64_set only work on an atomic size?
Really, I think we should just take a lock around both the read and the
write of dist->propbaser. It's not like this is going to be frequent or
in a critical path anywhere, right?
Thanks,
-Christoffer
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers
2016-08-02 14:55 ` Christoffer Dall
@ 2016-08-02 15:01 ` Marc Zyngier
0 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-08-02 15:01 UTC (permalink / raw)
To: linux-arm-kernel
On 02/08/16 15:55, Christoffer Dall wrote:
> On Tue, Aug 02, 2016 at 03:46:49PM +0100, Marc Zyngier wrote:
>> On 02/08/16 15:33, Christoffer Dall wrote:
>>> On Tue, Aug 02, 2016 at 11:12:46AM +0100, Marc Zyngier wrote:
>>>> On 02/08/16 10:40, Andre Przywara wrote:
>>>>> Hi,
>>>>>
>>>>> On 01/08/16 19:20, Christoffer Dall wrote:
>>>>>> On Fri, Jul 22, 2016 at 06:28:50PM +0100, Marc Zyngier wrote:
>>>>>>> From: Andre Przywara <andre.przywara@arm.com>
>>>>>>>
>>>>>>> In the GICv3 redistributor there are the PENDBASER and PROPBASER
>>>>>>> registers which we did not emulate so far, as they only make sense
>>>>>>> when having an ITS. In preparation for that emulate those MMIO
>>>>>>> accesses by storing the 64-bit data written into it into a variable
>>>>>>> which we later read in the ITS emulation.
>>>>>>> We also sanitise the registers, making sure RES0 regions are respected
>>>>>>> and checking for valid memory attributes.
>>>>>>>
>>>>>>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
>>>>>>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
>>>>>>> Tested-by: Eric Auger <eric.auger@redhat.com>
>>>>>>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
>>>>>>> ---
>>>>>>> include/kvm/arm_vgic.h | 13 ++++
>>>>>>> virt/kvm/arm/vgic/vgic-mmio-v3.c | 153 ++++++++++++++++++++++++++++++++++++++-
>>>>>>> virt/kvm/arm/vgic/vgic-mmio.h | 8 ++
>>>>>>> virt/kvm/arm/vgic/vgic-v3.c | 11 ++-
>>>>>>> 4 files changed, 181 insertions(+), 4 deletions(-)
>>>>>>>
>>>>>>> diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
>>>>>>> index 450b4da..df2dec5 100644
>>>>>>> --- a/include/kvm/arm_vgic.h
>>>>>>> +++ b/include/kvm/arm_vgic.h
>>>>>>> @@ -146,6 +146,14 @@ struct vgic_dist {
>>>>>>> struct vgic_irq *spis;
>>>>>>>
>>>>>>> struct vgic_io_device dist_iodev;
>>>>>>> +
>>>>>>> + /*
>>>>>>> + * Contains the attributes and gpa of the LPI configuration table.
>>>>>>> + * Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
>>>>>>> + * one address across all redistributors.
>>>>>>> + * GICv3 spec: 6.1.2 "LPI Configuration tables"
>>>>>>> + */
>>>>>>> + u64 propbaser;
>>>>>>> };
>>>>>>>
>>>>>>> struct vgic_v2_cpu_if {
>>>>>>> @@ -200,6 +208,11 @@ struct vgic_cpu {
>>>>>>> */
>>>>>>> struct vgic_io_device rd_iodev;
>>>>>>> struct vgic_io_device sgi_iodev;
>>>>>>> +
>>>>>>> + /* Contains the attributes and gpa of the LPI pending tables. */
>>>>>>> + u64 pendbaser;
>>>>>>> +
>>>>>>> + bool lpis_enabled;
>>>>>>> };
>>>>>>>
>>>>>>> int kvm_vgic_addr(struct kvm *kvm, unsigned long type, u64 *addr, bool write);
>>>>>>> diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>>>>>> index bfcafbd..278bfbb 100644
>>>>>>> --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>>>>>> +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
>>>>>>> @@ -29,6 +29,19 @@ static unsigned long extract_bytes(unsigned long data, unsigned int offset,
>>>>>>> return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
>>>>>>> }
>>>>>>>
>>>>>>> +/* allows updates of any half of a 64-bit register (or the whole thing) */
>>>>>>> +static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
>>>>>>> + unsigned long val)
>>>>>>> +{
>>>>>>> + int lower = (offset & 4) * 8;
>>>>>>> + int upper = lower + 8 * len - 1;
>>>>>>> +
>>>>>>> + reg &= ~GENMASK_ULL(upper, lower);
>>>>>>> + val &= GENMASK_ULL(len * 8 - 1, 0);
>>>>>>> +
>>>>>>> + return reg | ((u64)val << lower);
>>>>>>> +}
>>>>>>> +
>>>>>>> static unsined long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
>>>>>>> gpa_t addr, unsigned int len)
>>>>>>> {
>>>>>>> @@ -152,6 +165,142 @@ static unsigned long vgic_mmio_read_v3_idregs(struct kvm_vcpu *vcpu,
>>>>>>> return 0;
>>>>>>> }
>>>>>>>
>>>>>>> +/* We want to avoid outer shareable. */
>>>>>>> +u64 vgic_sanitise_shareability(u64 field)
>>>>>>> +{
>>>>>>> + switch (field) {
>>>>>>> + case GIC_BASER_OuterShareable:
>>>>>>> + return GIC_BASER_InnerShareable;
>>>>>>> + default:
>>>>>>> + return field;
>>>>>>> + }
>>>>>>> +}
>>>>>>> +
>>>>>>> +/* Avoid any inner non-cacheable mapping. */
>>>>>>> +u64 vgic_sanitise_inner_cacheability(u64 field)
>>>>>>> +{
>>>>>>> + switch (field) {
>>>>>>> + case GIC_BASER_CACHE_nCnB:
>>>>>>> + case GIC_BASER_CACHE_nC:
>>>>>>> + return GIC_BASER_CACHE_RaWb;
>>>>>>> + default:
>>>>>>> + return field;
>>>>>>> + }
>>>>>>> +}
>>>>>>> +
>>>>>>> +/* Non-cacheable or same-as-inner are OK. */
>>>>>>> +u64 vgic_sanitise_outer_cacheability(u64 field)
>>>>>>> +{
>>>>>>> + switch (field) {
>>>>>>> + case GIC_BASER_CACHE_SameAsInner:
>>>>>>> + case GIC_BASER_CACHE_nC:
>>>>>>> + return field;
>>>>>>> + default:
>>>>>>> + return GIC_BASER_CACHE_nC;
>>>>>>> + }
>>>>>>> +}
>>>>>>> +
>>>>>>> +u64 vgic_sanitise_field(u64 reg, u64 field_mask, int field_shift,
>>>>>>> + u64 (*sanitise_fn)(u64))
>>>>>>> +{
>>>>>>> + u64 field = (reg & field_mask) >> field_shift;
>>>>>>> +
>>>>>>> + field = sanitise_fn(field) << field_shift;
>>>>>>> + return (reg & ~field_mask) | field;
>>>>>>> +}
>>>>>>> +
>>>>>>> +#define PROPBASER_RES0_MASK \
>>>>>>> + (GENMASK_ULL(63, 59) | GENMASK_ULL(55, 52) | GENMASK_ULL(6, 5))
>>>>>>> +#define PENDBASER_RES0_MASK \
>>>>>>> + (BIT_ULL(63) | GENMASK_ULL(61, 59) | GENMASK_ULL(55, 52) | \
>>>>>>> + GENMASK_ULL(15, 12) | GENMASK_ULL(6, 0))
>>>>>>> +
>>>>>>> +static u64 vgic_sanitise_pendbaser(u64 reg)
>>>>>>> +{
>>>>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_SHAREABILITY_MASK,
>>>>>>> + GICR_PENDBASER_SHAREABILITY_SHIFT,
>>>>>>> + vgic_sanitise_shareability);
>>>>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_INNER_CACHEABILITY_MASK,
>>>>>>> + GICR_PENDBASER_INNER_CACHEABILITY_SHIFT,
>>>>>>> + vgic_sanitise_inner_cacheability);
>>>>>>> + reg = vgic_sanitise_field(reg, GICR_PENDBASER_OUTER_CACHEABILITY_MASK,
>>>>>>> + GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT,
>>>>>>> + vgic_sanitise_outer_cacheability);
>>>>>>> +
>>>>>>> + reg &= ~PENDBASER_RES0_MASK;
>>>>>>> + reg &= ~GENMASK_ULL(51, 48);
>>>>>>> +
>>>>>>> + return reg;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static u64 vgic_sanitise_propbaser(u64 reg)
>>>>>>> +{
>>>>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_SHAREABILITY_MASK,
>>>>>>> + GICR_PROPBASER_SHAREABILITY_SHIFT,
>>>>>>> + vgic_sanitise_shareability);
>>>>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_INNER_CACHEABILITY_MASK,
>>>>>>> + GICR_PROPBASER_INNER_CACHEABILITY_SHIFT,
>>>>>>> + vgic_sanitise_inner_cacheability);
>>>>>>> + reg = vgic_sanitise_field(reg, GICR_PROPBASER_OUTER_CACHEABILITY_MASK,
>>>>>>> + GICR_PROPBASER_OUTER_CACHEABILITY_SHIFT,
>>>>>>> + vgic_sanitise_outer_cacheability);
>>>>>>> +
>>>>>>> + reg &= ~PROPBASER_RES0_MASK;
>>>>>>> + reg &= ~GENMASK_ULL(51, 48);
>>>>>>> + return reg;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static unsigned long vgic_mmio_read_propbase(struct kvm_vcpu *vcpu,
>>>>>>> + gpa_t addr, unsigned int len)
>>>>>>> +{
>>>>>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>>>>>>> +
>>>>>>> + return extract_bytes(dist->propbaser, addr & 7, len);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void vgic_mmio_write_propbase(struct kvm_vcpu *vcpu,
>>>>>>> + gpa_t addr, unsigned int len,
>>>>>>> + unsigned long val)
>>>>>>> +{
>>>>>>> + struct vgic_dist *dist = &vcpu->kvm->arch.vgic;
>>>>>>> + struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
>>>>>>> + u64 propbaser = dist->propbaser;
>>>>>>> +
>>>>>>> + /* Storing a value with LPIs already enabled is undefined */
>>>>>>> + if (vgic_cpu->lpis_enabled)
>>>>>>> + return;
>>>>>>> +
>>>>>>> + propbaser = update_64bit_reg(propbaser, addr & 4, len, val);
>>>>>>> + propbaser = vgic_sanitise_propbaser(propbaser);
>>>>>>> +
>>>>>>> + dist->propbaser = propbaser;
>>>>>>
>>>>>> Which guarantees do we have that this will always be a single atomic
>>>>>> write?
>>>>>
>>>>> At least on arm64 - which is the only architecture this code is
>>>>> compiling for at the moment - I don't see why this shouldn't be a single
>>>>> write:
>>>>> str x19, [x22,#3544]
>>>>> is what my setup here creates.
>>>>>
>>>>> Do we need something stronger? Do we want to postpone this to the point
>>>>> when we get arm(32) support?
>>>>
>>>> This is a *device*. It shouldn't be affected by whatever drives it.
>>>>
>>>> Now, and more to the point: the write shouldn't have to be atomic. All
>>>> 64bit registers should be able to cope with 32bit writes to it, as
>>>> described in the architecture spec (IHI0069C 8.1.3).
>>>
>>> That's not my concern. My concern is that you have two CPUs updating
>>> the propbaser, once after the other, and you end up with a mix of the
>>> two updates. If this C-code can ever become two 32-bit writes, for
>>> example, and these can happen in parallel you can end up with something
>>> like that.
>>>
>>> If there is not valid expectation from guest software that a real device
>>> stores either one or the other value, then it's fine to leave it as is.
>>>
>>> Thinking about it, any sane guest would probably synchronize multiple
>>> writes to this register itself?
>>
>> Ah, I finally clocked what you mean. Yes, we need to guarantee that the
>> update to the register itself is atomic (architectural requirement).
>> Which means we need to turn this in to an atomic access (atomic64_set).
>
> Doesn't atomic64_set only work on an atomic size?
That would of course imply defining propbaser as an atomic64_t.
> Really, I think we should just take a lock around both the read and the
> write of dist->propbaser. It's not like this is going to be frequent or
> in a critical path anywhere, right?
That'd work as well for the MMIO side, though we shouldn't have to
evaluate it anywhere else as long as LPI are not enabled (and at which
point it is not able to change anymore).
Thanks,
M.
--
Jazz is not dead. It just smells funny...
^ permalink raw reply [flat|nested] 67+ messages in thread
* [PATCH 34/55] KVM: arm64: vgic-its: Introduce ITS emulation file with MMIO framework
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (32 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 33/55] KVM: arm64: vgic: Handle ITS related GICv3 redistributor registers Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 35/55] KVM: arm64: vgic-its: Introduce new KVM ITS device Marc Zyngier
` (21 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
The ARM GICv3 ITS emulation code goes into a separate file, but needs
to be connected to the GICv3 emulation, of which it is an option.
The ITS MMIO handlers require the respective ITS pointer to be passed in,
so we amend the existing VGIC MMIO framework to let it cope with that.
Also we introduce the basic ITS data structure and initialize it, but
don't return any success yet, as we are not yet ready for the show.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 22 ++++++++-
virt/kvm/arm/vgic/vgic-its.c | 103 +++++++++++++++++++++++++++++++++++++++
virt/kvm/arm/vgic/vgic-mmio-v3.c | 40 ++++++++++++++-
virt/kvm/arm/vgic/vgic-mmio.c | 37 +++++++++++---
virt/kvm/arm/vgic/vgic-mmio.h | 17 +++++--
virt/kvm/arm/vgic/vgic.h | 7 +++
6 files changed, 213 insertions(+), 13 deletions(-)
create mode 100644 virt/kvm/arm/vgic/vgic-its.c
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index df2dec5..685f339 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -108,15 +108,35 @@ struct vgic_irq {
};
struct vgic_register_region;
+struct vgic_its;
+
+enum iodev_type {
+ IODEV_CPUIF,
+ IODEV_DIST,
+ IODEV_REDIST,
+ IODEV_ITS
+};
struct vgic_io_device {
gpa_t base_addr;
- struct kvm_vcpu *redist_vcpu;
+ union {
+ struct kvm_vcpu *redist_vcpu;
+ struct vgic_its *its;
+ };
const struct vgic_register_region *regions;
+ enum iodev_type iodev_type;
int nr_regions;
struct kvm_io_device dev;
};
+struct vgic_its {
+ /* The base address of the ITS control register frame */
+ gpa_t vgic_its_base;
+
+ bool enabled;
+ struct vgic_io_device iodev;
+};
+
struct vgic_dist {
bool in_kernel;
bool ready;
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
new file mode 100644
index 0000000..4654d6e
--- /dev/null
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -0,0 +1,103 @@
+/*
+ * GICv3 ITS emulation
+ *
+ * Copyright (C) 2015,2016 ARM Ltd.
+ * Author: Andre Przywara <andre.przywara@arm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/cpu.h>
+#include <linux/kvm.h>
+#include <linux/kvm_host.h>
+#include <linux/interrupt.h>
+
+#include <linux/irqchip/arm-gic-v3.h>
+
+#include <asm/kvm_emulate.h>
+#include <asm/kvm_arm.h>
+#include <asm/kvm_mmu.h>
+
+#include "vgic.h"
+#include "vgic-mmio.h"
+
+#define REGISTER_ITS_DESC(off, rd, wr, length, acc) \
+{ \
+ .reg_offset = off, \
+ .len = length, \
+ .access_flags = acc, \
+ .its_read = rd, \
+ .its_write = wr, \
+}
+
+static unsigned long its_mmio_read_raz(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ return 0;
+}
+
+static void its_mmio_write_wi(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len, unsigned long val)
+{
+ /* Ignore */
+}
+
+static struct vgic_register_region its_registers[] = {
+ REGISTER_ITS_DESC(GITS_CTLR,
+ its_mmio_read_raz, its_mmio_write_wi, 4,
+ VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_IIDR,
+ its_mmio_read_raz, its_mmio_write_wi, 4,
+ VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_TYPER,
+ its_mmio_read_raz, its_mmio_write_wi, 8,
+ VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_CBASER,
+ its_mmio_read_raz, its_mmio_write_wi, 8,
+ VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_CWRITER,
+ its_mmio_read_raz, its_mmio_write_wi, 8,
+ VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_CREADR,
+ its_mmio_read_raz, its_mmio_write_wi, 8,
+ VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_BASER,
+ its_mmio_read_raz, its_mmio_write_wi, 0x40,
+ VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
+ REGISTER_ITS_DESC(GITS_IDREGS_BASE,
+ its_mmio_read_raz, its_mmio_write_wi, 0x30,
+ VGIC_ACCESS_32bit),
+};
+
+static int vgic_its_init_its(struct kvm *kvm, struct vgic_its *its)
+{
+ struct vgic_io_device *iodev = &its->iodev;
+ int ret;
+
+ if (IS_VGIC_ADDR_UNDEF(its->vgic_its_base))
+ return -ENXIO;
+
+ iodev->regions = its_registers;
+ iodev->nr_regions = ARRAY_SIZE(its_registers);
+ kvm_iodevice_init(&iodev->dev, &kvm_io_gic_ops);
+
+ iodev->base_addr = its->vgic_its_base;
+ iodev->iodev_type = IODEV_ITS;
+ iodev->its = its;
+ mutex_lock(&kvm->slots_lock);
+ ret = kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, iodev->base_addr,
+ KVM_VGIC_V3_ITS_SIZE, &iodev->dev);
+ mutex_unlock(&kvm->slots_lock);
+
+ return ret;
+}
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index 278bfbb..b92b7d6 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -42,6 +42,16 @@ static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
return reg | ((u64)val << lower);
}
+bool vgic_has_its(struct kvm *kvm)
+{
+ struct vgic_dist *dist = &kvm->arch.vgic;
+
+ if (dist->vgic_model != KVM_DEV_TYPE_ARM_VGIC_V3)
+ return false;
+
+ return false;
+}
+
static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
gpa_t addr, unsigned int len)
{
@@ -132,6 +142,32 @@ static void vgic_mmio_write_irouter(struct kvm_vcpu *vcpu,
vgic_put_irq(vcpu->kvm, irq);
}
+static unsigned long vgic_mmio_read_v3r_ctlr(struct kvm_vcpu *vcpu,
+ gpa_t addr, unsigned int len)
+{
+ struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+
+ return vgic_cpu->lpis_enabled ? GICR_CTLR_ENABLE_LPIS : 0;
+}
+
+
+static void vgic_mmio_write_v3r_ctlr(struct kvm_vcpu *vcpu,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+ bool was_enabled = vgic_cpu->lpis_enabled;
+
+ if (!vgic_has_its(vcpu->kvm))
+ return;
+
+ vgic_cpu->lpis_enabled = val & GICR_CTLR_ENABLE_LPIS;
+
+ if (!was_enabled && vgic_cpu->lpis_enabled) {
+ /* Eventually do something */
+ }
+}
+
static unsigned long vgic_mmio_read_v3r_typer(struct kvm_vcpu *vcpu,
gpa_t addr, unsigned int len)
{
@@ -372,7 +408,7 @@ static const struct vgic_register_region vgic_v3_dist_registers[] = {
static const struct vgic_register_region vgic_v3_rdbase_registers[] = {
REGISTER_DESC_WITH_LENGTH(GICR_CTLR,
- vgic_mmio_read_raz, vgic_mmio_write_wi, 4,
+ vgic_mmio_read_v3r_ctlr, vgic_mmio_write_v3r_ctlr, 4,
VGIC_ACCESS_32bit),
REGISTER_DESC_WITH_LENGTH(GICR_IIDR,
vgic_mmio_read_v3r_iidr, vgic_mmio_write_wi, 4,
@@ -450,6 +486,7 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
kvm_iodevice_init(&rd_dev->dev, &kvm_io_gic_ops);
rd_dev->base_addr = rd_base;
+ rd_dev->iodev_type = IODEV_REDIST;
rd_dev->regions = vgic_v3_rdbase_registers;
rd_dev->nr_regions = ARRAY_SIZE(vgic_v3_rdbase_registers);
rd_dev->redist_vcpu = vcpu;
@@ -464,6 +501,7 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t redist_base_address)
kvm_iodevice_init(&sgi_dev->dev, &kvm_io_gic_ops);
sgi_dev->base_addr = sgi_base;
+ sgi_dev->iodev_type = IODEV_REDIST;
sgi_dev->regions = vgic_v3_sgibase_registers;
sgi_dev->nr_regions = ARRAY_SIZE(vgic_v3_sgibase_registers);
sgi_dev->redist_vcpu = vcpu;
diff --git a/virt/kvm/arm/vgic/vgic-mmio.c b/virt/kvm/arm/vgic/vgic-mmio.c
index 5e79e01..26be827 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.c
+++ b/virt/kvm/arm/vgic/vgic-mmio.c
@@ -473,8 +473,7 @@ static int dispatch_mmio_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
{
struct vgic_io_device *iodev = kvm_to_vgic_iodev(dev);
const struct vgic_register_region *region;
- struct kvm_vcpu *r_vcpu;
- unsigned long data;
+ unsigned long data = 0;
region = vgic_find_mmio_region(iodev->regions, iodev->nr_regions,
addr - iodev->base_addr);
@@ -483,8 +482,20 @@ static int dispatch_mmio_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
return 0;
}
- r_vcpu = iodev->redist_vcpu ? iodev->redist_vcpu : vcpu;
- data = region->read(r_vcpu, addr, len);
+ switch (iodev->iodev_type) {
+ case IODEV_CPUIF:
+ return 1;
+ case IODEV_DIST:
+ data = region->read(vcpu, addr, len);
+ break;
+ case IODEV_REDIST:
+ data = region->read(iodev->redist_vcpu, addr, len);
+ break;
+ case IODEV_ITS:
+ data = region->its_read(vcpu->kvm, iodev->its, addr, len);
+ break;
+ }
+
vgic_data_host_to_mmio_bus(val, len, data);
return 0;
}
@@ -494,7 +505,6 @@ static int dispatch_mmio_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
{
struct vgic_io_device *iodev = kvm_to_vgic_iodev(dev);
const struct vgic_register_region *region;
- struct kvm_vcpu *r_vcpu;
unsigned long data = vgic_data_mmio_bus_to_host(val, len);
region = vgic_find_mmio_region(iodev->regions, iodev->nr_regions,
@@ -505,8 +515,20 @@ static int dispatch_mmio_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
if (!check_region(region, addr, len))
return 0;
- r_vcpu = iodev->redist_vcpu ? iodev->redist_vcpu : vcpu;
- region->write(r_vcpu, addr, len, data);
+ switch (iodev->iodev_type) {
+ case IODEV_CPUIF:
+ break;
+ case IODEV_DIST:
+ region->write(vcpu, addr, len, data);
+ break;
+ case IODEV_REDIST:
+ region->write(iodev->redist_vcpu, addr, len, data);
+ break;
+ case IODEV_ITS:
+ region->its_write(vcpu->kvm, iodev->its, addr, len, data);
+ break;
+ }
+
return 0;
}
@@ -536,6 +558,7 @@ int vgic_register_dist_iodev(struct kvm *kvm, gpa_t dist_base_address,
}
io_device->base_addr = dist_base_address;
+ io_device->iodev_type = IODEV_DIST;
io_device->redist_vcpu = NULL;
mutex_lock(&kvm->slots_lock);
diff --git a/virt/kvm/arm/vgic/vgic-mmio.h b/virt/kvm/arm/vgic/vgic-mmio.h
index 71aa39d..366d663 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.h
+++ b/virt/kvm/arm/vgic/vgic-mmio.h
@@ -21,10 +21,19 @@ struct vgic_register_region {
unsigned int len;
unsigned int bits_per_irq;
unsigned int access_flags;
- unsigned long (*read)(struct kvm_vcpu *vcpu, gpa_t addr,
- unsigned int len);
- void (*write)(struct kvm_vcpu *vcpu, gpa_t addr, unsigned int len,
- unsigned long val);
+ union {
+ unsigned long (*read)(struct kvm_vcpu *vcpu, gpa_t addr,
+ unsigned int len);
+ unsigned long (*its_read)(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len);
+ };
+ union {
+ void (*write)(struct kvm_vcpu *vcpu, gpa_t addr,
+ unsigned int len, unsigned long val);
+ void (*its_write)(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len,
+ unsigned long val);
+ };
};
extern struct kvm_io_device_ops kvm_io_gic_ops;
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 5b79c34..31807c1 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -72,6 +72,7 @@ void vgic_v3_enable(struct kvm_vcpu *vcpu);
int vgic_v3_probe(const struct gic_kvm_info *info);
int vgic_v3_map_resources(struct kvm *kvm);
int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t dist_base_address);
+bool vgic_has_its(struct kvm *kvm);
#else
static inline void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu)
{
@@ -123,6 +124,12 @@ static inline int vgic_register_redist_iodevs(struct kvm *kvm,
{
return -ENODEV;
}
+
+static inline bool vgic_has_its(struct kvm *kvm)
+{
+ return false;
+}
+
#endif
int kvm_register_vgic_device(unsigned long type);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 35/55] KVM: arm64: vgic-its: Introduce new KVM ITS device
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (33 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 34/55] KVM: arm64: vgic-its: Introduce ITS emulation file with MMIO framework Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 36/55] KVM: arm64: vgic-its: Implement basic ITS register handlers Marc Zyngier
` (20 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
Introduce a new KVM device that represents an ARM Interrupt Translation
Service (ITS) controller. Since there can be multiple of this per guest,
we can't piggy back on the existing GICv3 distributor device, but create
a new type of KVM device.
On the KVM_CREATE_DEVICE ioctl we allocate and initialize the ITS data
structure and store the pointer in the kvm_device data.
Upon an explicit init ioctl from userland (after having setup the MMIO
address) we register the handlers with the kvm_io_bus framework.
Any reference to an ITS thus has to go via this interface.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
Documentation/virtual/kvm/devices/arm-vgic.txt | 25 +++--
arch/arm/kvm/arm.c | 1 +
arch/arm64/include/uapi/asm/kvm.h | 2 +
include/kvm/arm_vgic.h | 3 +
include/uapi/linux/kvm.h | 2 +
virt/kvm/arm/vgic/vgic-its.c | 135 +++++++++++++++++++++++++
virt/kvm/arm/vgic/vgic-kvm-device.c | 4 +-
virt/kvm/arm/vgic/vgic-mmio-v3.c | 2 +-
virt/kvm/arm/vgic/vgic.h | 3 +
9 files changed, 168 insertions(+), 9 deletions(-)
diff --git a/Documentation/virtual/kvm/devices/arm-vgic.txt b/Documentation/virtual/kvm/devices/arm-vgic.txt
index 59541d4..89182f8 100644
--- a/Documentation/virtual/kvm/devices/arm-vgic.txt
+++ b/Documentation/virtual/kvm/devices/arm-vgic.txt
@@ -4,16 +4,22 @@ ARM Virtual Generic Interrupt Controller (VGIC)
Device types supported:
KVM_DEV_TYPE_ARM_VGIC_V2 ARM Generic Interrupt Controller v2.0
KVM_DEV_TYPE_ARM_VGIC_V3 ARM Generic Interrupt Controller v3.0
+ KVM_DEV_TYPE_ARM_VGIC_ITS ARM Interrupt Translation Service Controller
-Only one VGIC instance may be instantiated through either this API or the
-legacy KVM_CREATE_IRQCHIP api. The created VGIC will act as the VM interrupt
-controller, requiring emulated user-space devices to inject interrupts to the
-VGIC instead of directly to CPUs.
+Only one VGIC instance of the V2/V3 types above may be instantiated through
+either this API or the legacy KVM_CREATE_IRQCHIP api. The created VGIC will
+act as the VM interrupt controller, requiring emulated user-space devices to
+inject interrupts to the VGIC instead of directly to CPUs.
Creating a guest GICv3 device requires a host GICv3 as well.
GICv3 implementations with hardware compatibility support allow a guest GICv2
as well.
+Creating a virtual ITS controller requires a host GICv3 (but does not depend
+on having physical ITS controllers).
+There can be multiple ITS controllers per guest, each of them has to have
+a separate, non-overlapping MMIO region.
+
Groups:
KVM_DEV_ARM_VGIC_GRP_ADDR
Attributes:
@@ -39,6 +45,13 @@ Groups:
Only valid for KVM_DEV_TYPE_ARM_VGIC_V3.
This address needs to be 64K aligned.
+ KVM_VGIC_V3_ADDR_TYPE_ITS (rw, 64-bit)
+ Base address in the guest physical address space of the GICv3 ITS
+ control register frame. The ITS allows MSI(-X) interrupts to be
+ injected into guests. This extension is optional. If the kernel
+ does not support the ITS, the call returns -ENODEV.
+ Only valid for KVM_DEV_TYPE_ARM_VGIC_ITS.
+ This address needs to be 64K aligned and the region covers 128K.
KVM_DEV_ARM_VGIC_GRP_DIST_REGS
Attributes:
@@ -109,8 +122,8 @@ Groups:
KVM_DEV_ARM_VGIC_GRP_CTRL
Attributes:
KVM_DEV_ARM_VGIC_CTRL_INIT
- request the initialization of the VGIC, no additional parameter in
- kvm_device_attr.addr.
+ request the initialization of the VGIC or ITS, no additional parameter
+ in kvm_device_attr.addr.
Errors:
-ENXIO: VGIC not properly configured as required prior to calling
this attribute
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 972075c..fb4661c 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -20,6 +20,7 @@
#include <linux/errno.h>
#include <linux/err.h>
#include <linux/kvm_host.h>
+#include <linux/list.h>
#include <linux/module.h>
#include <linux/vmalloc.h>
#include <linux/fs.h>
diff --git a/arch/arm64/include/uapi/asm/kvm.h b/arch/arm64/include/uapi/asm/kvm.h
index f209ea1..3051f86 100644
--- a/arch/arm64/include/uapi/asm/kvm.h
+++ b/arch/arm64/include/uapi/asm/kvm.h
@@ -87,9 +87,11 @@ struct kvm_regs {
/* Supported VGICv3 address types */
#define KVM_VGIC_V3_ADDR_TYPE_DIST 2
#define KVM_VGIC_V3_ADDR_TYPE_REDIST 3
+#define KVM_VGIC_ITS_ADDR_TYPE 4
#define KVM_VGIC_V3_DIST_SIZE SZ_64K
#define KVM_VGIC_V3_REDIST_SIZE (2 * SZ_64K)
+#define KVM_VGIC_V3_ITS_SIZE (2 * SZ_64K)
#define KVM_ARM_VCPU_POWER_OFF 0 /* CPU is started in OFF state */
#define KVM_ARM_VCPU_EL1_32BIT 1 /* CPU running a 32bit VM */
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 685f339..8609fac 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -134,6 +134,7 @@ struct vgic_its {
gpa_t vgic_its_base;
bool enabled;
+ bool initialized;
struct vgic_io_device iodev;
};
@@ -167,6 +168,8 @@ struct vgic_dist {
struct vgic_io_device dist_iodev;
+ bool has_its;
+
/*
* Contains the attributes and gpa of the LPI configuration table.
* Since we report GICR_TYPER.CommonLPIAff as 0b00, we can share
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index 7de96f5..d8c4c32 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -1077,6 +1077,8 @@ enum kvm_device_type {
#define KVM_DEV_TYPE_FLIC KVM_DEV_TYPE_FLIC
KVM_DEV_TYPE_ARM_VGIC_V3,
#define KVM_DEV_TYPE_ARM_VGIC_V3 KVM_DEV_TYPE_ARM_VGIC_V3
+ KVM_DEV_TYPE_ARM_VGIC_ITS,
+#define KVM_DEV_TYPE_ARM_VGIC_ITS KVM_DEV_TYPE_ARM_VGIC_ITS
KVM_DEV_TYPE_MAX,
};
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 4654d6e..6b47b36 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -21,6 +21,7 @@
#include <linux/kvm.h>
#include <linux/kvm_host.h>
#include <linux/interrupt.h>
+#include <linux/uaccess.h>
#include <linux/irqchip/arm-gic-v3.h>
@@ -84,6 +85,9 @@ static int vgic_its_init_its(struct kvm *kvm, struct vgic_its *its)
struct vgic_io_device *iodev = &its->iodev;
int ret;
+ if (its->initialized)
+ return 0;
+
if (IS_VGIC_ADDR_UNDEF(its->vgic_its_base))
return -ENXIO;
@@ -99,5 +103,136 @@ static int vgic_its_init_its(struct kvm *kvm, struct vgic_its *its)
KVM_VGIC_V3_ITS_SIZE, &iodev->dev);
mutex_unlock(&kvm->slots_lock);
+ if (!ret)
+ its->initialized = true;
+
return ret;
}
+
+static int vgic_its_create(struct kvm_device *dev, u32 type)
+{
+ struct vgic_its *its;
+
+ if (type != KVM_DEV_TYPE_ARM_VGIC_ITS)
+ return -ENODEV;
+
+ its = kzalloc(sizeof(struct vgic_its), GFP_KERNEL);
+ if (!its)
+ return -ENOMEM;
+
+ its->vgic_its_base = VGIC_ADDR_UNDEF;
+
+ dev->kvm->arch.vgic.has_its = true;
+ its->initialized = false;
+ its->enabled = false;
+
+ dev->private = its;
+
+ return 0;
+}
+
+static void vgic_its_destroy(struct kvm_device *kvm_dev)
+{
+ struct vgic_its *its = kvm_dev->private;
+
+ kfree(its);
+}
+
+static int vgic_its_has_attr(struct kvm_device *dev,
+ struct kvm_device_attr *attr)
+{
+ switch (attr->group) {
+ case KVM_DEV_ARM_VGIC_GRP_ADDR:
+ switch (attr->attr) {
+ case KVM_VGIC_ITS_ADDR_TYPE:
+ return 0;
+ }
+ break;
+ case KVM_DEV_ARM_VGIC_GRP_CTRL:
+ switch (attr->attr) {
+ case KVM_DEV_ARM_VGIC_CTRL_INIT:
+ return 0;
+ }
+ break;
+ }
+ return -ENXIO;
+}
+
+static int vgic_its_set_attr(struct kvm_device *dev,
+ struct kvm_device_attr *attr)
+{
+ struct vgic_its *its = dev->private;
+ int ret;
+
+ switch (attr->group) {
+ case KVM_DEV_ARM_VGIC_GRP_ADDR: {
+ u64 __user *uaddr = (u64 __user *)(long)attr->addr;
+ unsigned long type = (unsigned long)attr->attr;
+ u64 addr;
+
+ if (type != KVM_VGIC_ITS_ADDR_TYPE)
+ return -ENODEV;
+
+ if (its->initialized)
+ return -EBUSY;
+
+ if (copy_from_user(&addr, uaddr, sizeof(addr)))
+ return -EFAULT;
+
+ ret = vgic_check_ioaddr(dev->kvm, &its->vgic_its_base,
+ addr, SZ_64K);
+ if (ret)
+ return ret;
+
+ its->vgic_its_base = addr;
+
+ return 0;
+ }
+ case KVM_DEV_ARM_VGIC_GRP_CTRL:
+ switch (attr->attr) {
+ case KVM_DEV_ARM_VGIC_CTRL_INIT:
+ return vgic_its_init_its(dev->kvm, its);
+ }
+ break;
+ }
+ return -ENXIO;
+}
+
+static int vgic_its_get_attr(struct kvm_device *dev,
+ struct kvm_device_attr *attr)
+{
+ switch (attr->group) {
+ case KVM_DEV_ARM_VGIC_GRP_ADDR: {
+ struct vgic_its *its = dev->private;
+ u64 addr = its->vgic_its_base;
+ u64 __user *uaddr = (u64 __user *)(long)attr->addr;
+ unsigned long type = (unsigned long)attr->attr;
+
+ if (type != KVM_VGIC_ITS_ADDR_TYPE)
+ return -ENODEV;
+
+ if (copy_to_user(uaddr, &addr, sizeof(addr)))
+ return -EFAULT;
+ break;
+ default:
+ return -ENXIO;
+ }
+ }
+
+ return 0;
+}
+
+static struct kvm_device_ops kvm_arm_vgic_its_ops = {
+ .name = "kvm-arm-vgic-its",
+ .create = vgic_its_create,
+ .destroy = vgic_its_destroy,
+ .set_attr = vgic_its_set_attr,
+ .get_attr = vgic_its_get_attr,
+ .has_attr = vgic_its_has_attr,
+};
+
+int kvm_vgic_register_its_device(void)
+{
+ return kvm_register_device_ops(&kvm_arm_vgic_its_ops,
+ KVM_DEV_TYPE_ARM_VGIC_ITS);
+}
diff --git a/virt/kvm/arm/vgic/vgic-kvm-device.c b/virt/kvm/arm/vgic/vgic-kvm-device.c
index 2f24f13..561d2ba 100644
--- a/virt/kvm/arm/vgic/vgic-kvm-device.c
+++ b/virt/kvm/arm/vgic/vgic-kvm-device.c
@@ -21,8 +21,8 @@
/* common helpers */
-static int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
- phys_addr_t addr, phys_addr_t alignment)
+int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
+ phys_addr_t addr, phys_addr_t alignment)
{
if (addr & ~KVM_PHYS_MASK)
return -E2BIG;
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index b92b7d6..a5c3505 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -49,7 +49,7 @@ bool vgic_has_its(struct kvm *kvm)
if (dist->vgic_model != KVM_DEV_TYPE_ARM_VGIC_V3)
return false;
- return false;
+ return dist->has_its;
}
static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 31807c1..8192a29 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -42,6 +42,9 @@ void vgic_put_irq(struct kvm *kvm, struct vgic_irq *irq);
bool vgic_queue_irq_unlock(struct kvm *kvm, struct vgic_irq *irq);
void vgic_kick_vcpus(struct kvm *kvm);
+int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
+ phys_addr_t addr, phys_addr_t alignment);
+
void vgic_v2_process_maintenance(struct kvm_vcpu *vcpu);
void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu);
void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 36/55] KVM: arm64: vgic-its: Implement basic ITS register handlers
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (34 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 35/55] KVM: arm64: vgic-its: Introduce new KVM ITS device Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 37/55] KVM: arm64: vgic-its: Connect LPIs to the VGIC emulation Marc Zyngier
` (19 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
Add emulation for some basic MMIO registers used in the ITS emulation.
This includes:
- GITS_{CTLR,TYPER,IIDR}
- ID registers
- GITS_{CBASER,CREADR,CWRITER}
(which implement the ITS command buffer handling)
- GITS_BASER<n>
Most of the handlers are pretty straight forward, only the CWRITER
handler is a bit more involved by taking the new its_cmd mutex and
then iterating over the command buffer.
The registers holding base addresses and attributes are sanitised before
storing them.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 16 ++
virt/kvm/arm/vgic/vgic-its.c | 399 +++++++++++++++++++++++++++++++++++++--
virt/kvm/arm/vgic/vgic-mmio-v3.c | 8 +-
virt/kvm/arm/vgic/vgic-mmio.h | 6 +
virt/kvm/arm/vgic/vgic.c | 12 +-
5 files changed, 420 insertions(+), 21 deletions(-)
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 8609fac..6186749 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -22,6 +22,7 @@
#include <linux/spinlock.h>
#include <linux/types.h>
#include <kvm/iodev.h>
+#include <linux/list.h>
#define VGIC_V3_MAX_CPUS 255
#define VGIC_V2_MAX_CPUS 8
@@ -136,6 +137,21 @@ struct vgic_its {
bool enabled;
bool initialized;
struct vgic_io_device iodev;
+
+ /* These registers correspond to GITS_BASER{0,1} */
+ u64 baser_device_table;
+ u64 baser_coll_table;
+
+ /* Protects the command queue */
+ struct mutex cmd_lock;
+ u64 cbaser;
+ u32 creadr;
+ u32 cwriter;
+
+ /* Protects the device and collection lists */
+ struct mutex its_lock;
+ struct list_head device_list;
+ struct list_head collection_list;
};
struct vgic_dist {
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 6b47b36..11cfe2f 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -21,6 +21,7 @@
#include <linux/kvm.h>
#include <linux/kvm_host.h>
#include <linux/interrupt.h>
+#include <linux/list.h>
#include <linux/uaccess.h>
#include <linux/irqchip/arm-gic-v3.h>
@@ -32,6 +33,329 @@
#include "vgic.h"
#include "vgic-mmio.h"
+struct its_device {
+ struct list_head dev_list;
+
+ /* the head for the list of ITTEs */
+ struct list_head itt_head;
+ u32 device_id;
+};
+
+#define COLLECTION_NOT_MAPPED ((u32)~0)
+
+struct its_collection {
+ struct list_head coll_list;
+
+ u32 collection_id;
+ u32 target_addr;
+};
+
+#define its_is_collection_mapped(coll) ((coll) && \
+ ((coll)->target_addr != COLLECTION_NOT_MAPPED))
+
+struct its_itte {
+ struct list_head itte_list;
+
+ struct its_collection *collection;
+ u32 lpi;
+ u32 event_id;
+};
+
+/*
+ * We only implement 48 bits of PA at the moment, although the ITS
+ * supports more. Let's be restrictive here.
+ */
+#define CBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 12))
+
+static unsigned long vgic_mmio_read_its_ctlr(struct kvm *vcpu,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ u32 reg = 0;
+
+ mutex_lock(&its->cmd_lock);
+ if (its->creadr == its->cwriter)
+ reg |= GITS_CTLR_QUIESCENT;
+ if (its->enabled)
+ reg |= GITS_CTLR_ENABLE;
+ mutex_unlock(&its->cmd_lock);
+
+ return reg;
+}
+
+static void vgic_mmio_write_its_ctlr(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ its->enabled = !!(val & GITS_CTLR_ENABLE);
+}
+
+static unsigned long vgic_mmio_read_its_typer(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ u64 reg = GITS_TYPER_PLPIS;
+
+ /*
+ * We use linear CPU numbers for redistributor addressing,
+ * so GITS_TYPER.PTA is 0.
+ * Also we force all PROPBASER registers to be the same, so
+ * CommonLPIAff is 0 as well.
+ * To avoid memory waste in the guest, we keep the number of IDBits and
+ * DevBits low - as least for the time being.
+ */
+ reg |= 0x0f << GITS_TYPER_DEVBITS_SHIFT;
+ reg |= 0x0f << GITS_TYPER_IDBITS_SHIFT;
+
+ return extract_bytes(reg, addr & 7, len);
+}
+
+static unsigned long vgic_mmio_read_its_iidr(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ return (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
+}
+
+static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ switch (addr & 0xffff) {
+ case GITS_PIDR0:
+ return 0x92; /* part number, bits[7:0] */
+ case GITS_PIDR1:
+ return 0xb4; /* part number, bits[11:8] */
+ case GITS_PIDR2:
+ return GIC_PIDR2_ARCH_GICv3 | 0x0b;
+ case GITS_PIDR4:
+ return 0x40; /* This is a 64K software visible page */
+ /* The following are the ID registers for (any) GIC. */
+ case GITS_CIDR0:
+ return 0x0d;
+ case GITS_CIDR1:
+ return 0xf0;
+ case GITS_CIDR2:
+ return 0x05;
+ case GITS_CIDR3:
+ return 0xb1;
+ }
+
+ return 0;
+}
+
+/* Requires the its_lock to be held. */
+static void its_free_itte(struct kvm *kvm, struct its_itte *itte)
+{
+ list_del(&itte->itte_list);
+ kfree(itte);
+}
+
+static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ return -ENODEV;
+}
+
+static u64 vgic_sanitise_its_baser(u64 reg)
+{
+ reg = vgic_sanitise_field(reg, GITS_BASER_SHAREABILITY_MASK,
+ GITS_BASER_SHAREABILITY_SHIFT,
+ vgic_sanitise_shareability);
+ reg = vgic_sanitise_field(reg, GITS_BASER_INNER_CACHEABILITY_MASK,
+ GITS_BASER_INNER_CACHEABILITY_SHIFT,
+ vgic_sanitise_inner_cacheability);
+ reg = vgic_sanitise_field(reg, GITS_BASER_OUTER_CACHEABILITY_MASK,
+ GITS_BASER_OUTER_CACHEABILITY_SHIFT,
+ vgic_sanitise_outer_cacheability);
+
+ /* Bits 15:12 contain bits 51:48 of the PA, which we don't support. */
+ reg &= ~GENMASK_ULL(15, 12);
+
+ /* We support only one (ITS) page size: 64K */
+ reg = (reg & ~GITS_BASER_PAGE_SIZE_MASK) | GITS_BASER_PAGE_SIZE_64K;
+
+ return reg;
+}
+
+static u64 vgic_sanitise_its_cbaser(u64 reg)
+{
+ reg = vgic_sanitise_field(reg, GITS_CBASER_SHAREABILITY_MASK,
+ GITS_CBASER_SHAREABILITY_SHIFT,
+ vgic_sanitise_shareability);
+ reg = vgic_sanitise_field(reg, GITS_CBASER_INNER_CACHEABILITY_MASK,
+ GITS_CBASER_INNER_CACHEABILITY_SHIFT,
+ vgic_sanitise_inner_cacheability);
+ reg = vgic_sanitise_field(reg, GITS_CBASER_OUTER_CACHEABILITY_MASK,
+ GITS_CBASER_OUTER_CACHEABILITY_SHIFT,
+ vgic_sanitise_outer_cacheability);
+
+ /*
+ * Sanitise the physical address to be 64k aligned.
+ * Also limit the physical addresses to 48 bits.
+ */
+ reg &= ~(GENMASK_ULL(51, 48) | GENMASK_ULL(15, 12));
+
+ return reg;
+}
+
+static unsigned long vgic_mmio_read_its_cbaser(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ return extract_bytes(its->cbaser, addr & 7, len);
+}
+
+static void vgic_mmio_write_its_cbaser(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ /* When GITS_CTLR.Enable is 1, this register is RO. */
+ if (its->enabled)
+ return;
+
+ mutex_lock(&its->cmd_lock);
+ its->cbaser = update_64bit_reg(its->cbaser, addr & 7, len, val);
+ its->cbaser = vgic_sanitise_its_cbaser(its->cbaser);
+ its->creadr = 0;
+ /*
+ * CWRITER is architecturally UNKNOWN on reset, but we need to reset
+ * it to CREADR to make sure we start with an empty command buffer.
+ */
+ its->cwriter = its->creadr;
+ mutex_unlock(&its->cmd_lock);
+}
+
+#define ITS_CMD_BUFFER_SIZE(baser) ((((baser) & 0xff) + 1) << 12)
+#define ITS_CMD_SIZE 32
+#define ITS_CMD_OFFSET(reg) ((reg) & GENMASK(19, 5))
+
+/*
+ * By writing to CWRITER the guest announces new commands to be processed.
+ * To avoid any races in the first place, we take the its_cmd lock, which
+ * protects our ring buffer variables, so that there is only one user
+ * per ITS handling commands@a given time.
+ */
+static void vgic_mmio_write_its_cwriter(struct kvm *kvm, struct vgic_its *its,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ gpa_t cbaser;
+ u64 cmd_buf[4];
+ u32 reg;
+
+ if (!its)
+ return;
+
+ mutex_lock(&its->cmd_lock);
+
+ reg = update_64bit_reg(its->cwriter, addr & 7, len, val);
+ reg = ITS_CMD_OFFSET(reg);
+ if (reg >= ITS_CMD_BUFFER_SIZE(its->cbaser)) {
+ mutex_unlock(&its->cmd_lock);
+ return;
+ }
+
+ its->cwriter = reg;
+ cbaser = CBASER_ADDRESS(its->cbaser);
+
+ while (its->cwriter != its->creadr) {
+ int ret = kvm_read_guest(kvm, cbaser + its->creadr,
+ cmd_buf, ITS_CMD_SIZE);
+ /*
+ * If kvm_read_guest() fails, this could be due to the guest
+ * programming a bogus value in CBASER or something else going
+ * wrong from which we cannot easily recover.
+ * According to section 6.3.2 in the GICv3 spec we can just
+ * ignore that command then.
+ */
+ if (!ret)
+ vgic_its_handle_command(kvm, its, cmd_buf);
+
+ its->creadr += ITS_CMD_SIZE;
+ if (its->creadr == ITS_CMD_BUFFER_SIZE(its->cbaser))
+ its->creadr = 0;
+ }
+
+ mutex_unlock(&its->cmd_lock);
+}
+
+static unsigned long vgic_mmio_read_its_cwriter(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ return extract_bytes(its->cwriter, addr & 0x7, len);
+}
+
+static unsigned long vgic_mmio_read_its_creadr(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ return extract_bytes(its->creadr, addr & 0x7, len);
+}
+
+#define BASER_INDEX(addr) (((addr) / sizeof(u64)) & 0x7)
+static unsigned long vgic_mmio_read_its_baser(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len)
+{
+ u64 reg;
+
+ switch (BASER_INDEX(addr)) {
+ case 0:
+ reg = its->baser_device_table;
+ break;
+ case 1:
+ reg = its->baser_coll_table;
+ break;
+ default:
+ reg = 0;
+ break;
+ }
+
+ return extract_bytes(reg, addr & 7, len);
+}
+
+#define GITS_BASER_RO_MASK (GENMASK_ULL(52, 48) | GENMASK_ULL(58, 56))
+static void vgic_mmio_write_its_baser(struct kvm *kvm,
+ struct vgic_its *its,
+ gpa_t addr, unsigned int len,
+ unsigned long val)
+{
+ u64 entry_size, device_type;
+ u64 reg, *regptr, clearbits = 0;
+
+ /* When GITS_CTLR.Enable is 1, we ignore write accesses. */
+ if (its->enabled)
+ return;
+
+ switch (BASER_INDEX(addr)) {
+ case 0:
+ regptr = &its->baser_device_table;
+ entry_size = 8;
+ device_type = GITS_BASER_TYPE_DEVICE;
+ break;
+ case 1:
+ regptr = &its->baser_coll_table;
+ entry_size = 8;
+ device_type = GITS_BASER_TYPE_COLLECTION;
+ clearbits = GITS_BASER_INDIRECT;
+ break;
+ default:
+ return;
+ }
+
+ reg = update_64bit_reg(*regptr, addr & 7, len, val);
+ reg &= ~GITS_BASER_RO_MASK;
+ reg &= ~clearbits;
+
+ reg |= (entry_size - 1) << GITS_BASER_ENTRY_SIZE_SHIFT;
+ reg |= device_type << GITS_BASER_TYPE_SHIFT;
+ reg = vgic_sanitise_its_baser(reg);
+
+ *regptr = reg;
+}
+
#define REGISTER_ITS_DESC(off, rd, wr, length, acc) \
{ \
.reg_offset = off, \
@@ -41,12 +365,6 @@
.its_write = wr, \
}
-static unsigned long its_mmio_read_raz(struct kvm *kvm, struct vgic_its *its,
- gpa_t addr, unsigned int len)
-{
- return 0;
-}
-
static void its_mmio_write_wi(struct kvm *kvm, struct vgic_its *its,
gpa_t addr, unsigned int len, unsigned long val)
{
@@ -55,28 +373,28 @@ static void its_mmio_write_wi(struct kvm *kvm, struct vgic_its *its,
static struct vgic_register_region its_registers[] = {
REGISTER_ITS_DESC(GITS_CTLR,
- its_mmio_read_raz, its_mmio_write_wi, 4,
+ vgic_mmio_read_its_ctlr, vgic_mmio_write_its_ctlr, 4,
VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_IIDR,
- its_mmio_read_raz, its_mmio_write_wi, 4,
+ vgic_mmio_read_its_iidr, its_mmio_write_wi, 4,
VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_TYPER,
- its_mmio_read_raz, its_mmio_write_wi, 8,
+ vgic_mmio_read_its_typer, its_mmio_write_wi, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_CBASER,
- its_mmio_read_raz, its_mmio_write_wi, 8,
+ vgic_mmio_read_its_cbaser, vgic_mmio_write_its_cbaser, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_CWRITER,
- its_mmio_read_raz, its_mmio_write_wi, 8,
+ vgic_mmio_read_its_cwriter, vgic_mmio_write_its_cwriter, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_CREADR,
- its_mmio_read_raz, its_mmio_write_wi, 8,
+ vgic_mmio_read_its_creadr, its_mmio_write_wi, 8,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_BASER,
- its_mmio_read_raz, its_mmio_write_wi, 0x40,
+ vgic_mmio_read_its_baser, vgic_mmio_write_its_baser, 0x40,
VGIC_ACCESS_64bit | VGIC_ACCESS_32bit),
REGISTER_ITS_DESC(GITS_IDREGS_BASE,
- its_mmio_read_raz, its_mmio_write_wi, 0x30,
+ vgic_mmio_read_its_idregs, its_mmio_write_wi, 0x30,
VGIC_ACCESS_32bit),
};
@@ -109,6 +427,18 @@ static int vgic_its_init_its(struct kvm *kvm, struct vgic_its *its)
return ret;
}
+#define INITIAL_BASER_VALUE \
+ (GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWb) | \
+ GIC_BASER_CACHEABILITY(GITS_BASER, OUTER, SameAsInner) | \
+ GIC_BASER_SHAREABILITY(GITS_BASER, InnerShareable) | \
+ ((8ULL - 1) << GITS_BASER_ENTRY_SIZE_SHIFT) | \
+ GITS_BASER_PAGE_SIZE_64K)
+
+#define INITIAL_PROPBASER_VALUE \
+ (GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWb) | \
+ GIC_BASER_CACHEABILITY(GICR_PROPBASER, OUTER, SameAsInner) | \
+ GIC_BASER_SHAREABILITY(GICR_PROPBASER, InnerShareable))
+
static int vgic_its_create(struct kvm_device *dev, u32 type)
{
struct vgic_its *its;
@@ -120,12 +450,24 @@ static int vgic_its_create(struct kvm_device *dev, u32 type)
if (!its)
return -ENOMEM;
+ mutex_init(&its->its_lock);
+ mutex_init(&its->cmd_lock);
+
its->vgic_its_base = VGIC_ADDR_UNDEF;
+ INIT_LIST_HEAD(&its->device_list);
+ INIT_LIST_HEAD(&its->collection_list);
+
dev->kvm->arch.vgic.has_its = true;
its->initialized = false;
its->enabled = false;
+ its->baser_device_table = INITIAL_BASER_VALUE |
+ ((u64)GITS_BASER_TYPE_DEVICE << GITS_BASER_TYPE_SHIFT);
+ its->baser_coll_table = INITIAL_BASER_VALUE |
+ ((u64)GITS_BASER_TYPE_COLLECTION << GITS_BASER_TYPE_SHIFT);
+ dev->kvm->arch.vgic.propbaser = INITIAL_PROPBASER_VALUE;
+
dev->private = its;
return 0;
@@ -133,7 +475,36 @@ static int vgic_its_create(struct kvm_device *dev, u32 type)
static void vgic_its_destroy(struct kvm_device *kvm_dev)
{
+ struct kvm *kvm = kvm_dev->kvm;
struct vgic_its *its = kvm_dev->private;
+ struct its_device *dev;
+ struct its_itte *itte;
+ struct list_head *dev_cur, *dev_temp;
+ struct list_head *cur, *temp;
+
+ /*
+ * We may end up here without the lists ever having been initialized.
+ * Check this and bail out early to avoid dereferencing a NULL pointer.
+ */
+ if (!its->device_list.next)
+ return;
+
+ mutex_lock(&its->its_lock);
+ list_for_each_safe(dev_cur, dev_temp, &its->device_list) {
+ dev = container_of(dev_cur, struct its_device, dev_list);
+ list_for_each_safe(cur, temp, &dev->itt_head) {
+ itte = (container_of(cur, struct its_itte, itte_list));
+ its_free_itte(kvm, itte);
+ }
+ list_del(dev_cur);
+ kfree(dev);
+ }
+
+ list_for_each_safe(cur, temp, &its->collection_list) {
+ list_del(cur);
+ kfree(container_of(cur, struct its_collection, coll_list));
+ }
+ mutex_unlock(&its->its_lock);
kfree(its);
}
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index a5c3505..84a301d 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -23,15 +23,15 @@
#include "vgic-mmio.h"
/* extract @num bytes at @offset bytes offset in data */
-static unsigned long extract_bytes(unsigned long data, unsigned int offset,
- unsigned int num)
+unsigned long extract_bytes(unsigned long data, unsigned int offset,
+ unsigned int num)
{
return (data >> (offset * 8)) & GENMASK_ULL(num * 8 - 1, 0);
}
/* allows updates of any half of a 64-bit register (or the whole thing) */
-static u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
- unsigned long val)
+u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
+ unsigned long val)
{
int lower = (offset & 4) * 8;
int upper = lower + 8 * len - 1;
diff --git a/virt/kvm/arm/vgic/vgic-mmio.h b/virt/kvm/arm/vgic/vgic-mmio.h
index 366d663..0b3ecf9 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.h
+++ b/virt/kvm/arm/vgic/vgic-mmio.h
@@ -96,6 +96,12 @@ unsigned long vgic_data_mmio_bus_to_host(const void *val, unsigned int len);
void vgic_data_host_to_mmio_bus(void *buf, unsigned int len,
unsigned long data);
+unsigned long extract_bytes(unsigned long data, unsigned int offset,
+ unsigned int num);
+
+u64 update_64bit_reg(u64 reg, unsigned int offset, unsigned int len,
+ unsigned long val);
+
unsigned long vgic_mmio_read_raz(struct kvm_vcpu *vcpu,
gpa_t addr, unsigned int len);
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index fb19a55..d3ba1b4 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -33,10 +33,16 @@ struct vgic_global __section(.hyp.text) kvm_vgic_global_state;
/*
* Locking order is always:
- * vgic_cpu->ap_list_lock
- * vgic_irq->irq_lock
+ * its->cmd_lock (mutex)
+ * its->its_lock (mutex)
+ * vgic_cpu->ap_list_lock
+ * vgic_irq->irq_lock
*
- * (that is, always take the ap_list_lock before the struct vgic_irq lock).
+ * If you need to take multiple locks, always take the upper lock first,
+ * then the lower ones, e.g. first take the its_lock, then the irq_lock.
+ * If you are already holding a lock and need to take a higher one, you
+ * have to drop the lower ranking lock first and re-aquire it after having
+ * taken the upper one.
*
* When taking more than one ap_list_lock at the same time, always take the
* lowest numbered VCPU's ap_list_lock first, so:
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 37/55] KVM: arm64: vgic-its: Connect LPIs to the VGIC emulation
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (35 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 36/55] KVM: arm64: vgic-its: Implement basic ITS register handlers Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 38/55] KVM: arm64: vgic-its: Read initial LPI pending table Marc Zyngier
` (18 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
LPIs are dynamically created (mapped) at guest runtime and their
actual number can be quite high, but is mostly assigned using a very
sparse allocation scheme. So arrays are not an ideal data structure
to hold the information.
We use a spin-lock protected linked list to hold all mapped LPIs,
represented by their struct vgic_irq. This lock is grouped between the
ap_list_lock and the vgic_irq lock in our locking order.
Also we store a pointer to that struct vgic_irq in our struct its_itte,
so we can easily access it.
Eventually we call our new vgic_get_lpi() from vgic_get_irq(), so
the VGIC code gets transparently access to LPIs.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 6 +++++
virt/kvm/arm/vgic/vgic-init.c | 3 +++
virt/kvm/arm/vgic/vgic-its.c | 5 ++++
virt/kvm/arm/vgic/vgic-v3.c | 2 ++
virt/kvm/arm/vgic/vgic.c | 63 ++++++++++++++++++++++++++++++++++++++-----
5 files changed, 73 insertions(+), 6 deletions(-)
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 6186749..a6ca326 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -77,6 +77,7 @@ enum vgic_irq_config {
struct vgic_irq {
spinlock_t irq_lock; /* Protects the content of the struct */
+ struct list_head lpi_list; /* Used to link all LPIs together */
struct list_head ap_list;
struct kvm_vcpu *vcpu; /* SGIs and PPIs: The VCPU
@@ -193,6 +194,11 @@ struct vgic_dist {
* GICv3 spec: 6.1.2 "LPI Configuration tables"
*/
u64 propbaser;
+
+ /* Protects the lpi_list and the count value below. */
+ spinlock_t lpi_list_lock;
+ struct list_head lpi_list_head;
+ int lpi_list_count;
};
struct vgic_v2_cpu_if {
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index ac3c1a5..535e713 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -157,6 +157,9 @@ static int kvm_vgic_dist_init(struct kvm *kvm, unsigned int nr_spis)
struct kvm_vcpu *vcpu0 = kvm_get_vcpu(kvm, 0);
int i;
+ INIT_LIST_HEAD(&dist->lpi_list_head);
+ spin_lock_init(&dist->lpi_list_lock);
+
dist->spis = kcalloc(nr_spis, sizeof(struct vgic_irq), GFP_KERNEL);
if (!dist->spis)
return -ENOMEM;
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 11cfe2f..14f91ff 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -56,6 +56,7 @@ struct its_collection {
struct its_itte {
struct list_head itte_list;
+ struct vgic_irq *irq;
struct its_collection *collection;
u32 lpi;
u32 event_id;
@@ -148,6 +149,10 @@ static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
static void its_free_itte(struct kvm *kvm, struct its_itte *itte)
{
list_del(&itte->itte_list);
+
+ /* This put matches the get in vgic_add_lpi. */
+ vgic_put_irq(kvm, itte->irq);
+
kfree(itte);
}
diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c
index 6f8f31f..0506543 100644
--- a/virt/kvm/arm/vgic/vgic-v3.c
+++ b/virt/kvm/arm/vgic/vgic-v3.c
@@ -81,6 +81,8 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
else
intid = val & GICH_LR_VIRTUALID;
irq = vgic_get_irq(vcpu->kvm, vcpu, intid);
+ if (!irq) /* An LPI could have been unmapped. */
+ continue;
spin_lock(&irq->irq_lock);
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index d3ba1b4..53299fc 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -36,7 +36,8 @@ struct vgic_global __section(.hyp.text) kvm_vgic_global_state;
* its->cmd_lock (mutex)
* its->its_lock (mutex)
* vgic_cpu->ap_list_lock
- * vgic_irq->irq_lock
+ * kvm->lpi_list_lock
+ * vgic_irq->irq_lock
*
* If you need to take multiple locks, always take the upper lock first,
* then the lower ones, e.g. first take the its_lock, then the irq_lock.
@@ -51,6 +52,41 @@ struct vgic_global __section(.hyp.text) kvm_vgic_global_state;
* spin_lock(vcpuY->arch.vgic_cpu.ap_list_lock);
*/
+/*
+ * Iterate over the VM's list of mapped LPIs to find the one with a
+ * matching interrupt ID and return a reference to the IRQ structure.
+ */
+static struct vgic_irq *vgic_get_lpi(struct kvm *kvm, u32 intid)
+{
+ struct vgic_dist *dist = &kvm->arch.vgic;
+ struct vgic_irq *irq = NULL;
+
+ spin_lock(&dist->lpi_list_lock);
+
+ list_for_each_entry(irq, &dist->lpi_list_head, lpi_list) {
+ if (irq->intid != intid)
+ continue;
+
+ /*
+ * This increases the refcount, the caller is expected to
+ * call vgic_put_irq() later once it's finished with the IRQ.
+ */
+ kref_get(&irq->refcount);
+ goto out_unlock;
+ }
+ irq = NULL;
+
+out_unlock:
+ spin_unlock(&dist->lpi_list_lock);
+
+ return irq;
+}
+
+/*
+ * This looks up the virtual interrupt ID to get the corresponding
+ * struct vgic_irq. It also increases the refcount, so any caller is expected
+ * to call vgic_put_irq() once it's finished with this IRQ.
+ */
struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
u32 intid)
{
@@ -62,9 +98,9 @@ struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
if (intid <= VGIC_MAX_SPI)
return &kvm->arch.vgic.spis[intid - VGIC_NR_PRIVATE_IRQS];
- /* LPIs are not yet covered */
+ /* LPIs */
if (intid >= VGIC_MIN_LPI)
- return NULL;
+ return vgic_get_lpi(kvm, intid);
WARN(1, "Looking up struct vgic_irq for reserved INTID");
return NULL;
@@ -78,18 +114,33 @@ static void vgic_get_irq_kref(struct vgic_irq *irq)
kref_get(&irq->refcount);
}
-/* The refcount should never drop to 0 at the moment. */
+/*
+ * We can't do anything in here, because we lack the kvm pointer to
+ * lock and remove the item from the lpi_list. So we keep this function
+ * empty and use the return value of kref_put() to trigger the freeing.
+ */
static void vgic_irq_release(struct kref *ref)
{
- WARN_ON(1);
}
void vgic_put_irq(struct kvm *kvm, struct vgic_irq *irq)
{
+ struct vgic_dist *dist;
+
if (irq->intid < VGIC_MIN_LPI)
return;
- kref_put(&irq->refcount, vgic_irq_release);
+ if (!kref_put(&irq->refcount, vgic_irq_release))
+ return;
+
+ dist = &kvm->arch.vgic;
+
+ spin_lock(&dist->lpi_list_lock);
+ list_del(&irq->lpi_list);
+ dist->lpi_list_count--;
+ spin_unlock(&dist->lpi_list_lock);
+
+ kfree(irq);
}
/**
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 38/55] KVM: arm64: vgic-its: Read initial LPI pending table
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (36 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 37/55] KVM: arm64: vgic-its: Connect LPIs to the VGIC emulation Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 39/55] KVM: arm64: vgic-its: Allow updates of LPI configuration table Marc Zyngier
` (17 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
The LPI pending status for a GICv3 redistributor is held in a table
in (guest) memory. To achieve reasonable performance, we cache the
pending bit in our struct vgic_irq. The initial pending state must be
read from guest memory upon enabling LPIs for this redistributor.
As we can't access the guest memory while we hold the lpi_list spinlock,
we create a snapshot of the LPI list and iterate over that.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 95 ++++++++++++++++++++++++++++++++++++++++++++
virt/kvm/arm/vgic/vgic.h | 5 +++
2 files changed, 100 insertions(+)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 14f91ff..2881b84 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -67,6 +67,94 @@ struct its_itte {
* supports more. Let's be restrictive here.
*/
#define CBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 12))
+#define PENDBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 16))
+
+/*
+ * Create a snapshot of the current LPI list, so that we can enumerate all
+ * LPIs without holding any lock.
+ * Returns the array length and puts the kmalloc'ed array into intid_ptr.
+ */
+static int vgic_copy_lpi_list(struct kvm *kvm, u32 **intid_ptr)
+{
+ struct vgic_dist *dist = &kvm->arch.vgic;
+ struct vgic_irq *irq;
+ u32 *intids;
+ int irq_count = dist->lpi_list_count, i = 0;
+
+ /*
+ * We use the current value of the list length, which may change
+ * after the kmalloc. We don't care, because the guest shouldn't
+ * change anything while the command handling is still running,
+ * and in the worst case we would miss a new IRQ, which one wouldn't
+ * expect to be covered by this command anyway.
+ */
+ intids = kmalloc_array(irq_count, sizeof(intids[0]), GFP_KERNEL);
+ if (!intids)
+ return -ENOMEM;
+
+ spin_lock(&dist->lpi_list_lock);
+ list_for_each_entry(irq, &dist->lpi_list_head, lpi_list) {
+ /* We don't need to "get" the IRQ, as we hold the list lock. */
+ intids[i] = irq->intid;
+ if (++i == irq_count)
+ break;
+ }
+ spin_unlock(&dist->lpi_list_lock);
+
+ *intid_ptr = intids;
+ return irq_count;
+}
+
+/*
+ * Scan the whole LPI pending table and sync the pending bit in there
+ * with our own data structures. This relies on the LPI being
+ * mapped before.
+ */
+static int its_sync_lpi_pending_table(struct kvm_vcpu *vcpu)
+{
+ gpa_t pendbase = PENDBASER_ADDRESS(vcpu->arch.vgic_cpu.pendbaser);
+ struct vgic_irq *irq;
+ int last_byte_offset = -1;
+ int ret = 0;
+ u32 *intids;
+ int nr_irqs, i;
+
+ nr_irqs = vgic_copy_lpi_list(vcpu->kvm, &intids);
+ if (nr_irqs < 0)
+ return nr_irqs;
+
+ for (i = 0; i < nr_irqs; i++) {
+ int byte_offset, bit_nr;
+ u8 pendmask;
+
+ byte_offset = intids[i] / BITS_PER_BYTE;
+ bit_nr = intids[i] % BITS_PER_BYTE;
+
+ /*
+ * For contiguously allocated LPIs chances are we just read
+ * this very same byte in the last iteration. Reuse that.
+ */
+ if (byte_offset != last_byte_offset) {
+ ret = kvm_read_guest(vcpu->kvm, pendbase + byte_offset,
+ &pendmask, 1);
+ if (ret) {
+ kfree(intids);
+ return ret;
+ }
+ last_byte_offset = byte_offset;
+ }
+
+ irq = vgic_get_irq(vcpu->kvm, NULL, intids[i]);
+ spin_lock(&irq->irq_lock);
+ irq->pending = pendmask & (1U << bit_nr);
+ vgic_queue_irq_unlock(vcpu->kvm, irq);
+ vgic_put_irq(vcpu->kvm, irq);
+ }
+
+ kfree(intids);
+
+ return ret;
+}
static unsigned long vgic_mmio_read_its_ctlr(struct kvm *vcpu,
struct vgic_its *its,
@@ -403,6 +491,13 @@ static struct vgic_register_region its_registers[] = {
VGIC_ACCESS_32bit),
};
+/* This is called on setting the LPI enable bit in the redistributor. */
+void vgic_enable_lpis(struct kvm_vcpu *vcpu)
+{
+ if (!(vcpu->arch.vgic_cpu.pendbaser & GICR_PENDBASER_PTZ))
+ its_sync_lpi_pending_table(vcpu);
+}
+
static int vgic_its_init_its(struct kvm *kvm, struct vgic_its *its)
{
struct vgic_io_device *iodev = &its->iodev;
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 8192a29..ee348de 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -25,6 +25,7 @@
#define IS_VGIC_ADDR_UNDEF(_x) ((_x) == VGIC_ADDR_UNDEF)
#define INTERRUPT_ID_BITS_SPIS 10
+#define INTERRUPT_ID_BITS_ITS 16
#define VGIC_PRI_BITS 5
#define vgic_irq_is_sgi(intid) ((intid) < VGIC_NR_SGIS)
@@ -76,6 +77,7 @@ int vgic_v3_probe(const struct gic_kvm_info *info);
int vgic_v3_map_resources(struct kvm *kvm);
int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t dist_base_address);
bool vgic_has_its(struct kvm *kvm);
+void vgic_enable_lpis(struct kvm_vcpu *vcpu);
#else
static inline void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu)
{
@@ -133,6 +135,9 @@ static inline bool vgic_has_its(struct kvm *kvm)
return false;
}
+static inline void vgic_enable_lpis(struct kvm_vcpu *vcpu)
+{
+}
#endif
int kvm_register_vgic_device(unsigned long type);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 39/55] KVM: arm64: vgic-its: Allow updates of LPI configuration table
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (37 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 38/55] KVM: arm64: vgic-its: Read initial LPI pending table Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 40/55] KVM: arm64: vgic-its: Implement ITS command queue command handlers Marc Zyngier
` (16 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
The (system-wide) LPI configuration table is held in a table in
(guest) memory. To achieve reasonable performance, we cache this data
in our struct vgic_irq. If the guest updates the configuration data
(which consists of the enable bit and the priority value), it issues
an INV or INVALL command to allow us to update our information.
Provide functions that update that information for one LPI or all LPIs
mapped to a specific collection.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 39 +++++++++++++++++++++++++++++++++++++++
1 file changed, 39 insertions(+)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 2881b84..6f43b3b 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -68,6 +68,45 @@ struct its_itte {
*/
#define CBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 12))
#define PENDBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 16))
+#define PROPBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 12))
+
+#define GIC_LPI_OFFSET 8192
+
+#define LPI_PROP_ENABLE_BIT(p) ((p) & LPI_PROP_ENABLED)
+#define LPI_PROP_PRIORITY(p) ((p) & 0xfc)
+
+/*
+ * Reads the configuration data for a given LPI from guest memory and
+ * updates the fields in struct vgic_irq.
+ * If filter_vcpu is not NULL, applies only if the IRQ is targeting this
+ * VCPU. Unconditionally applies if filter_vcpu is NULL.
+ */
+static int update_lpi_config(struct kvm *kvm, struct vgic_irq *irq,
+ struct kvm_vcpu *filter_vcpu)
+{
+ u64 propbase = PROPBASER_ADDRESS(kvm->arch.vgic.propbaser);
+ u8 prop;
+ int ret;
+
+ ret = kvm_read_guest(kvm, propbase + irq->intid - GIC_LPI_OFFSET,
+ &prop, 1);
+
+ if (ret)
+ return ret;
+
+ spin_lock(&irq->irq_lock);
+
+ if (!filter_vcpu || filter_vcpu == irq->target_vcpu) {
+ irq->priority = LPI_PROP_PRIORITY(prop);
+ irq->enabled = LPI_PROP_ENABLE_BIT(prop);
+
+ vgic_queue_irq_unlock(kvm, irq);
+ } else {
+ spin_unlock(&irq->irq_lock);
+ }
+
+ return 0;
+}
/*
* Create a snapshot of the current LPI list, so that we can enumerate all
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 40/55] KVM: arm64: vgic-its: Implement ITS command queue command handlers
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (38 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 39/55] KVM: arm64: vgic-its: Allow updates of LPI configuration table Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:28 ` [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation Marc Zyngier
` (15 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
The connection between a device, an event ID, the LPI number and the
associated CPU is stored in in-memory tables in a GICv3, but their
format is not specified by the spec. Instead software uses a command
queue in a ring buffer to let an ITS implementation use its own
format.
Implement handlers for the various ITS commands and let them store
the requested relation into our own data structures. Those data
structures are protected by the its_lock mutex.
Our internal ring buffer read and write pointers are protected by the
its_cmd mutex, so that only one VCPU per ITS can handle commands at
any given time.
Error handling is very basic at the moment, as we don't have a good
way of communicating errors to the guest (usually an SError).
The INT command handler is missing from this patch, as we gain the
capability of actually injecting MSIs into the guest only later on.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 661 ++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 660 insertions(+), 1 deletion(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 6f43b3b..1408c88 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -33,6 +33,67 @@
#include "vgic.h"
#include "vgic-mmio.h"
+/*
+ * Creates a new (reference to a) struct vgic_irq for a given LPI.
+ * If this LPI is already mapped on another ITS, we increase its refcount
+ * and return a pointer to the existing structure.
+ * If this is a "new" LPI, we allocate and initialize a new struct vgic_irq.
+ * This function returns a pointer to the _unlocked_ structure.
+ */
+static struct vgic_irq *vgic_add_lpi(struct kvm *kvm, u32 intid)
+{
+ struct vgic_dist *dist = &kvm->arch.vgic;
+ struct vgic_irq *irq = vgic_get_irq(kvm, NULL, intid), *oldirq;
+
+ /* In this case there is no put, since we keep the reference. */
+ if (irq)
+ return irq;
+
+ irq = kzalloc(sizeof(struct vgic_irq), GFP_KERNEL);
+ if (!irq)
+ return NULL;
+
+ INIT_LIST_HEAD(&irq->lpi_list);
+ INIT_LIST_HEAD(&irq->ap_list);
+ spin_lock_init(&irq->irq_lock);
+
+ irq->config = VGIC_CONFIG_EDGE;
+ kref_init(&irq->refcount);
+ irq->intid = intid;
+
+ spin_lock(&dist->lpi_list_lock);
+
+ /*
+ * There could be a race with another vgic_add_lpi(), so we need to
+ * check that we don't add a second list entry with the same LPI.
+ */
+ list_for_each_entry(oldirq, &dist->lpi_list_head, lpi_list) {
+ if (oldirq->intid != intid)
+ continue;
+
+ /* Someone was faster with adding this LPI, lets use that. */
+ kfree(irq);
+ irq = oldirq;
+
+ /*
+ * This increases the refcount, the caller is expected to
+ * call vgic_put_irq() on the returned pointer once it's
+ * finished with the IRQ.
+ */
+ kref_get(&irq->refcount);
+
+ goto out_unlock;
+ }
+
+ list_add_tail(&irq->lpi_list, &dist->lpi_list_head);
+ dist->lpi_list_count++;
+
+out_unlock:
+ spin_unlock(&dist->lpi_list_lock);
+
+ return irq;
+}
+
struct its_device {
struct list_head dev_list;
@@ -63,15 +124,74 @@ struct its_itte {
};
/*
+ * Find and returns a device in the device table for an ITS.
+ * Must be called with the its_lock mutex held.
+ */
+static struct its_device *find_its_device(struct vgic_its *its, u32 device_id)
+{
+ struct its_device *device;
+
+ list_for_each_entry(device, &its->device_list, dev_list)
+ if (device_id == device->device_id)
+ return device;
+
+ return NULL;
+}
+
+/*
+ * Find and returns an interrupt translation table entry (ITTE) for a given
+ * Device ID/Event ID pair on an ITS.
+ * Must be called with the its_lock mutex held.
+ */
+static struct its_itte *find_itte(struct vgic_its *its, u32 device_id,
+ u32 event_id)
+{
+ struct its_device *device;
+ struct its_itte *itte;
+
+ device = find_its_device(its, device_id);
+ if (device == NULL)
+ return NULL;
+
+ list_for_each_entry(itte, &device->itt_head, itte_list)
+ if (itte->event_id == event_id)
+ return itte;
+
+ return NULL;
+}
+
+/* To be used as an iterator this macro misses the enclosing parentheses */
+#define for_each_lpi_its(dev, itte, its) \
+ list_for_each_entry(dev, &(its)->device_list, dev_list) \
+ list_for_each_entry(itte, &(dev)->itt_head, itte_list)
+
+/*
* We only implement 48 bits of PA at the moment, although the ITS
* supports more. Let's be restrictive here.
*/
+#define BASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 16))
#define CBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 12))
#define PENDBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 16))
#define PROPBASER_ADDRESS(x) ((x) & GENMASK_ULL(47, 12))
#define GIC_LPI_OFFSET 8192
+/*
+ * Finds and returns a collection in the ITS collection table.
+ * Must be called with the its_lock mutex held.
+ */
+static struct its_collection *find_collection(struct vgic_its *its, int coll_id)
+{
+ struct its_collection *collection;
+
+ list_for_each_entry(collection, &its->collection_list, coll_list) {
+ if (coll_id == collection->collection_id)
+ return collection;
+ }
+
+ return NULL;
+}
+
#define LPI_PROP_ENABLE_BIT(p) ((p) & LPI_PROP_ENABLED)
#define LPI_PROP_PRIORITY(p) ((p) & 0xfc)
@@ -145,6 +265,51 @@ static int vgic_copy_lpi_list(struct kvm *kvm, u32 **intid_ptr)
}
/*
+ * Promotes the ITS view of affinity of an ITTE (which redistributor this LPI
+ * is targeting) to the VGIC's view, which deals with target VCPUs.
+ * Needs to be called whenever either the collection for a LPIs has
+ * changed or the collection itself got retargeted.
+ */
+static void update_affinity_itte(struct kvm *kvm, struct its_itte *itte)
+{
+ struct kvm_vcpu *vcpu;
+
+ if (!its_is_collection_mapped(itte->collection))
+ return;
+
+ vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
+
+ spin_lock(&itte->irq->irq_lock);
+ itte->irq->target_vcpu = vcpu;
+ spin_unlock(&itte->irq->irq_lock);
+}
+
+/*
+ * Updates the target VCPU for every LPI targeting this collection.
+ * Must be called with the its_lock mutex held.
+ */
+static void update_affinity_collection(struct kvm *kvm, struct vgic_its *its,
+ struct its_collection *coll)
+{
+ struct its_device *device;
+ struct its_itte *itte;
+
+ for_each_lpi_its(device, itte, its) {
+ if (!itte->collection || coll != itte->collection)
+ continue;
+
+ update_affinity_itte(kvm, itte);
+ }
+}
+
+static u32 max_lpis_propbaser(u64 propbaser)
+{
+ int nr_idbits = (propbaser & 0x1f) + 1;
+
+ return 1U << min(nr_idbits, INTERRUPT_ID_BITS_ITS);
+}
+
+/*
* Scan the whole LPI pending table and sync the pending bit in there
* with our own data structures. This relies on the LPI being
* mapped before.
@@ -283,10 +448,504 @@ static void its_free_itte(struct kvm *kvm, struct its_itte *itte)
kfree(itte);
}
+static u64 its_cmd_mask_field(u64 *its_cmd, int word, int shift, int size)
+{
+ return (le64_to_cpu(its_cmd[word]) >> shift) & (BIT_ULL(size) - 1);
+}
+
+#define its_cmd_get_command(cmd) its_cmd_mask_field(cmd, 0, 0, 8)
+#define its_cmd_get_deviceid(cmd) its_cmd_mask_field(cmd, 0, 32, 32)
+#define its_cmd_get_id(cmd) its_cmd_mask_field(cmd, 1, 0, 32)
+#define its_cmd_get_physical_id(cmd) its_cmd_mask_field(cmd, 1, 32, 32)
+#define its_cmd_get_collection(cmd) its_cmd_mask_field(cmd, 2, 0, 16)
+#define its_cmd_get_target_addr(cmd) its_cmd_mask_field(cmd, 2, 16, 32)
+#define its_cmd_get_validbit(cmd) its_cmd_mask_field(cmd, 2, 63, 1)
+
+/*
+ * The DISCARD command frees an Interrupt Translation Table Entry (ITTE).
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_discard(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 device_id = its_cmd_get_deviceid(its_cmd);
+ u32 event_id = its_cmd_get_id(its_cmd);
+ struct its_itte *itte;
+
+
+ itte = find_itte(its, device_id, event_id);
+ if (itte && itte->collection) {
+ /*
+ * Though the spec talks about removing the pending state, we
+ * don't bother here since we clear the ITTE anyway and the
+ * pending state is a property of the ITTE struct.
+ */
+ its_free_itte(kvm, itte);
+ return 0;
+ }
+
+ return E_ITS_DISCARD_UNMAPPED_INTERRUPT;
+}
+
+/*
+ * The MOVI command moves an ITTE to a different collection.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_movi(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 device_id = its_cmd_get_deviceid(its_cmd);
+ u32 event_id = its_cmd_get_id(its_cmd);
+ u32 coll_id = its_cmd_get_collection(its_cmd);
+ struct kvm_vcpu *vcpu;
+ struct its_itte *itte;
+ struct its_collection *collection;
+
+ itte = find_itte(its, device_id, event_id);
+ if (!itte)
+ return E_ITS_MOVI_UNMAPPED_INTERRUPT;
+
+ if (!its_is_collection_mapped(itte->collection))
+ return E_ITS_MOVI_UNMAPPED_COLLECTION;
+
+ collection = find_collection(its, coll_id);
+ if (!its_is_collection_mapped(collection))
+ return E_ITS_MOVI_UNMAPPED_COLLECTION;
+
+ itte->collection = collection;
+ vcpu = kvm_get_vcpu(kvm, collection->target_addr);
+
+ spin_lock(&itte->irq->irq_lock);
+ itte->irq->target_vcpu = vcpu;
+ spin_unlock(&itte->irq->irq_lock);
+
+ return 0;
+}
+
+static void vgic_its_init_collection(struct vgic_its *its,
+ struct its_collection *collection,
+ u32 coll_id)
+{
+ collection->collection_id = coll_id;
+ collection->target_addr = COLLECTION_NOT_MAPPED;
+
+ list_add_tail(&collection->coll_list, &its->collection_list);
+}
+
+/*
+ * The MAPTI and MAPI commands map LPIs to ITTEs.
+ * Must be called with its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd, u8 subcmd)
+{
+ u32 device_id = its_cmd_get_deviceid(its_cmd);
+ u32 event_id = its_cmd_get_id(its_cmd);
+ u32 coll_id = its_cmd_get_collection(its_cmd);
+ struct its_itte *itte;
+ struct its_device *device;
+ struct its_collection *collection, *new_coll = NULL;
+ int lpi_nr;
+
+ device = find_its_device(its, device_id);
+ if (!device)
+ return E_ITS_MAPTI_UNMAPPED_DEVICE;
+
+ collection = find_collection(its, coll_id);
+ if (!collection) {
+ new_coll = kzalloc(sizeof(struct its_collection), GFP_KERNEL);
+ if (!new_coll)
+ return -ENOMEM;
+ }
+
+ if (subcmd == GITS_CMD_MAPTI)
+ lpi_nr = its_cmd_get_physical_id(its_cmd);
+ else
+ lpi_nr = event_id;
+ if (lpi_nr < GIC_LPI_OFFSET ||
+ lpi_nr >= max_lpis_propbaser(kvm->arch.vgic.propbaser)) {
+ kfree(new_coll);
+ return E_ITS_MAPTI_PHYSICALID_OOR;
+ }
+
+ itte = find_itte(its, device_id, event_id);
+ if (!itte) {
+ itte = kzalloc(sizeof(struct its_itte), GFP_KERNEL);
+ if (!itte) {
+ kfree(new_coll);
+ return -ENOMEM;
+ }
+
+ itte->event_id = event_id;
+ list_add_tail(&itte->itte_list, &device->itt_head);
+ }
+
+ if (!collection) {
+ collection = new_coll;
+ vgic_its_init_collection(its, collection, coll_id);
+ }
+
+ itte->collection = collection;
+ itte->lpi = lpi_nr;
+ itte->irq = vgic_add_lpi(kvm, lpi_nr);
+ update_affinity_itte(kvm, itte);
+
+ /*
+ * We "cache" the configuration table entries in out struct vgic_irq's.
+ * However we only have those structs for mapped IRQs, so we read in
+ * the respective config data from memory here upon mapping the LPI.
+ */
+ update_lpi_config(kvm, itte->irq, NULL);
+
+ return 0;
+}
+
+/* Requires the its_lock to be held. */
+static void vgic_its_unmap_device(struct kvm *kvm, struct its_device *device)
+{
+ struct its_itte *itte, *temp;
+
+ /*
+ * The spec says that unmapping a device with still valid
+ * ITTEs associated is UNPREDICTABLE. We remove all ITTEs,
+ * since we cannot leave the memory unreferenced.
+ */
+ list_for_each_entry_safe(itte, temp, &device->itt_head, itte_list)
+ its_free_itte(kvm, itte);
+
+ list_del(&device->dev_list);
+ kfree(device);
+}
+
+/*
+ * Check whether a device ID can be stored into the guest device tables.
+ * For a direct table this is pretty easy, but gets a bit nasty for
+ * indirect tables. We check whether the resulting guest physical address
+ * is actually valid (covered by a memslot and guest accessbible).
+ * For this we have to read the respective first level entry.
+ */
+static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
+ int device_id)
+{
+ u64 r = its->baser_device_table;
+ int nr_entries = GITS_BASER_NR_PAGES(r) * SZ_64K;
+ int index;
+ u64 indirect_ptr;
+ gfn_t gfn;
+
+
+ if (!(r & GITS_BASER_INDIRECT))
+ return device_id < (nr_entries / GITS_BASER_ENTRY_SIZE(r));
+
+ /* calculate and check the index into the 1st level */
+ index = device_id / (SZ_64K / GITS_BASER_ENTRY_SIZE(r));
+ if (index >= (nr_entries / sizeof(u64)))
+ return false;
+
+ /* Each 1st level entry is represented by a 64-bit value. */
+ if (!kvm_read_guest(kvm,
+ BASER_ADDRESS(r) + index * sizeof(indirect_ptr),
+ &indirect_ptr, sizeof(indirect_ptr)))
+ return false;
+
+ /* check the valid bit of the first level entry */
+ if (!(indirect_ptr & BIT_ULL(63)))
+ return false;
+
+ /*
+ * Mask the guest physical address and calculate the frame number.
+ * Any address beyond our supported 48 bits of PA will be caught
+ * by the actual check in the final step.
+ */
+ gfn = (indirect_ptr & GENMASK_ULL(51, 16)) >> PAGE_SHIFT;
+
+ return kvm_is_visible_gfn(kvm, gfn);
+}
+
+/*
+ * MAPD maps or unmaps a device ID to Interrupt Translation Tables (ITTs).
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_mapd(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 device_id = its_cmd_get_deviceid(its_cmd);
+ bool valid = its_cmd_get_validbit(its_cmd);
+ struct its_device *device;
+
+ if (!vgic_its_check_device_id(kvm, its, device_id))
+ return E_ITS_MAPD_DEVICE_OOR;
+
+ device = find_its_device(its, device_id);
+
+ /*
+ * The spec says that calling MAPD on an already mapped device
+ * invalidates all cached data for this device. We implement this
+ * by removing the mapping and re-establishing it.
+ */
+ if (device)
+ vgic_its_unmap_device(kvm, device);
+
+ /*
+ * The spec does not say whether unmapping a not-mapped device
+ * is an error, so we are done in any case.
+ */
+ if (!valid)
+ return 0;
+
+ device = kzalloc(sizeof(struct its_device), GFP_KERNEL);
+ if (!device)
+ return -ENOMEM;
+
+ device->device_id = device_id;
+ INIT_LIST_HEAD(&device->itt_head);
+
+ list_add_tail(&device->dev_list, &its->device_list);
+
+ return 0;
+}
+
+static int vgic_its_nr_collection_ids(struct vgic_its *its)
+{
+ u64 r = its->baser_coll_table;
+
+ return (GITS_BASER_NR_PAGES(r) * SZ_64K) / GITS_BASER_ENTRY_SIZE(r);
+}
+
+/*
+ * The MAPC command maps collection IDs to redistributors.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_mapc(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u16 coll_id;
+ u32 target_addr;
+ struct its_collection *collection;
+ bool valid;
+
+ valid = its_cmd_get_validbit(its_cmd);
+ coll_id = its_cmd_get_collection(its_cmd);
+ target_addr = its_cmd_get_target_addr(its_cmd);
+
+ if (target_addr >= atomic_read(&kvm->online_vcpus))
+ return E_ITS_MAPC_PROCNUM_OOR;
+
+ if (coll_id >= vgic_its_nr_collection_ids(its))
+ return E_ITS_MAPC_COLLECTION_OOR;
+
+ collection = find_collection(its, coll_id);
+
+ if (!valid) {
+ struct its_device *device;
+ struct its_itte *itte;
+ /*
+ * Clearing the mapping for that collection ID removes the
+ * entry from the list. If there wasn't any before, we can
+ * go home early.
+ */
+ if (!collection)
+ return 0;
+
+ for_each_lpi_its(device, itte, its)
+ if (itte->collection &&
+ itte->collection->collection_id == coll_id)
+ itte->collection = NULL;
+
+ list_del(&collection->coll_list);
+ kfree(collection);
+ } else {
+ if (!collection) {
+ collection = kzalloc(sizeof(struct its_collection),
+ GFP_KERNEL);
+ if (!collection)
+ return -ENOMEM;
+
+ vgic_its_init_collection(its, collection, coll_id);
+ collection->target_addr = target_addr;
+ } else {
+ collection->target_addr = target_addr;
+ update_affinity_collection(kvm, its, collection);
+ }
+ }
+
+ return 0;
+}
+
+/*
+ * The CLEAR command removes the pending state for a particular LPI.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_clear(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 device_id = its_cmd_get_deviceid(its_cmd);
+ u32 event_id = its_cmd_get_id(its_cmd);
+ struct its_itte *itte;
+
+
+ itte = find_itte(its, device_id, event_id);
+ if (!itte)
+ return E_ITS_CLEAR_UNMAPPED_INTERRUPT;
+
+ itte->irq->pending = false;
+
+ return 0;
+}
+
+/*
+ * The INV command syncs the configuration bits from the memory table.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_inv(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 device_id = its_cmd_get_deviceid(its_cmd);
+ u32 event_id = its_cmd_get_id(its_cmd);
+ struct its_itte *itte;
+
+
+ itte = find_itte(its, device_id, event_id);
+ if (!itte)
+ return E_ITS_INV_UNMAPPED_INTERRUPT;
+
+ return update_lpi_config(kvm, itte->irq, NULL);
+}
+
+/*
+ * The INVALL command requests flushing of all IRQ data in this collection.
+ * Find the VCPU mapped to that collection, then iterate over the VM's list
+ * of mapped LPIs and update the configuration for each IRQ which targets
+ * the specified vcpu. The configuration will be read from the in-memory
+ * configuration table.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_invall(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 coll_id = its_cmd_get_collection(its_cmd);
+ struct its_collection *collection;
+ struct kvm_vcpu *vcpu;
+ struct vgic_irq *irq;
+ u32 *intids;
+ int irq_count, i;
+
+ collection = find_collection(its, coll_id);
+ if (!its_is_collection_mapped(collection))
+ return E_ITS_INVALL_UNMAPPED_COLLECTION;
+
+ vcpu = kvm_get_vcpu(kvm, collection->target_addr);
+
+ irq_count = vgic_copy_lpi_list(kvm, &intids);
+ if (irq_count < 0)
+ return irq_count;
+
+ for (i = 0; i < irq_count; i++) {
+ irq = vgic_get_irq(kvm, NULL, intids[i]);
+ if (!irq)
+ continue;
+ update_lpi_config(kvm, irq, vcpu);
+ vgic_put_irq(kvm, irq);
+ }
+
+ kfree(intids);
+
+ return 0;
+}
+
+/*
+ * The MOVALL command moves the pending state of all IRQs targeting one
+ * redistributor to another. We don't hold the pending state in the VCPUs,
+ * but in the IRQs instead, so there is really not much to do for us here.
+ * However the spec says that no IRQ must target the old redistributor
+ * afterwards, so we make sure that no LPI is using the associated target_vcpu.
+ * This command affects all LPIs in the system that target that redistributor.
+ */
+static int vgic_its_cmd_handle_movall(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ struct vgic_dist *dist = &kvm->arch.vgic;
+ u32 target1_addr = its_cmd_get_target_addr(its_cmd);
+ u32 target2_addr = its_cmd_mask_field(its_cmd, 3, 16, 32);
+ struct kvm_vcpu *vcpu1, *vcpu2;
+ struct vgic_irq *irq;
+
+ if (target1_addr >= atomic_read(&kvm->online_vcpus) ||
+ target2_addr >= atomic_read(&kvm->online_vcpus))
+ return E_ITS_MOVALL_PROCNUM_OOR;
+
+ if (target1_addr == target2_addr)
+ return 0;
+
+ vcpu1 = kvm_get_vcpu(kvm, target1_addr);
+ vcpu2 = kvm_get_vcpu(kvm, target2_addr);
+
+ spin_lock(&dist->lpi_list_lock);
+
+ list_for_each_entry(irq, &dist->lpi_list_head, lpi_list) {
+ spin_lock(&irq->irq_lock);
+
+ if (irq->target_vcpu == vcpu1)
+ irq->target_vcpu = vcpu2;
+
+ spin_unlock(&irq->irq_lock);
+ }
+
+ spin_unlock(&dist->lpi_list_lock);
+
+ return 0;
+}
+
+/*
+ * This function is called with the its_cmd lock held, but the ITS data
+ * structure lock dropped.
+ */
static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
u64 *its_cmd)
{
- return -ENODEV;
+ u8 cmd = its_cmd_get_command(its_cmd);
+ int ret = -ENODEV;
+
+ mutex_lock(&its->its_lock);
+ switch (cmd) {
+ case GITS_CMD_MAPD:
+ ret = vgic_its_cmd_handle_mapd(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_MAPC:
+ ret = vgic_its_cmd_handle_mapc(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_MAPI:
+ ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd, cmd);
+ break;
+ case GITS_CMD_MAPTI:
+ ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd, cmd);
+ break;
+ case GITS_CMD_MOVI:
+ ret = vgic_its_cmd_handle_movi(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_DISCARD:
+ ret = vgic_its_cmd_handle_discard(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_CLEAR:
+ ret = vgic_its_cmd_handle_clear(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_MOVALL:
+ ret = vgic_its_cmd_handle_movall(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_INV:
+ ret = vgic_its_cmd_handle_inv(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_INVALL:
+ ret = vgic_its_cmd_handle_invall(kvm, its, its_cmd);
+ break;
+ case GITS_CMD_SYNC:
+ /* we ignore this command: we are in sync all of the time */
+ ret = 0;
+ break;
+ }
+ mutex_unlock(&its->its_lock);
+
+ return ret;
}
static u64 vgic_sanitise_its_baser(u64 reg)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (39 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 40/55] KVM: arm64: vgic-its: Implement ITS command queue command handlers Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-08-01 18:20 ` Christoffer Dall
2016-08-04 10:47 ` Christoffer Dall
2016-07-22 17:28 ` [PATCH 42/55] KVM: arm64: vgic-its: Enable ITS emulation as a virtual MSI controller Marc Zyngier
` (14 subsequent siblings)
55 siblings, 2 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
When userland wants to inject an MSI into the guest, it uses the
KVM_SIGNAL_MSI ioctl, which carries the doorbell address along with
the payload and the device ID.
With the help of the KVM IO bus framework we learn the corresponding
ITS from the doorbell address. We then use our wrapper functions to
iterate the linked lists and find the proper Interrupt Translation Table
Entry (ITTE) and thus the corresponding struct vgic_irq to finally set
the pending bit.
We also provide the handler for the ITS "INT" command, which allows a
guest to trigger an MSI via the ITS command queue. Since this one knows
about the right ITS already, we directly call the MMIO handler function
without using the kvm_io_bus framework.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 77 ++++++++++++++++++++++++++++++++++++++++++++
virt/kvm/arm/vgic/vgic.h | 6 ++++
2 files changed, 83 insertions(+)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 1408c88..d8e8f14 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -437,6 +437,65 @@ static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
return 0;
}
+/*
+ * Find the target VCPU and the LPI number for a given devid/eventid pair
+ * and make this IRQ pending, possibly injecting it.
+ * Must be called with the its_lock mutex held.
+ */
+static void vgic_its_trigger_msi(struct kvm *kvm, struct vgic_its *its,
+ u32 devid, u32 eventid)
+{
+ struct its_itte *itte;
+
+ if (!its->enabled)
+ return;
+
+ itte = find_itte(its, devid, eventid);
+ /* Triggering an unmapped IRQ gets silently dropped. */
+ if (itte && its_is_collection_mapped(itte->collection)) {
+ struct kvm_vcpu *vcpu;
+
+ vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
+ if (vcpu && vcpu->arch.vgic_cpu.lpis_enabled) {
+ spin_lock(&itte->irq->irq_lock);
+ itte->irq->pending = true;
+ vgic_queue_irq_unlock(kvm, itte->irq);
+ }
+ }
+}
+
+/*
+ * Queries the KVM IO bus framework to get the ITS pointer from the given
+ * doorbell address.
+ * We then call vgic_its_trigger_msi() with the decoded data.
+ */
+int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
+{
+ u64 address;
+ struct kvm_io_device *kvm_io_dev;
+ struct vgic_io_device *iodev;
+
+ if (!vgic_has_its(kvm))
+ return -ENODEV;
+
+ if (!(msi->flags & KVM_MSI_VALID_DEVID))
+ return -EINVAL;
+
+ address = (u64)msi->address_hi << 32 | msi->address_lo;
+
+ kvm_io_dev = kvm_io_bus_get_dev(kvm, KVM_MMIO_BUS, address);
+ if (!kvm_io_dev)
+ return -ENODEV;
+
+ iodev = container_of(kvm_io_dev, struct vgic_io_device, dev);
+
+ mutex_lock(&iodev->its->its_lock);
+ vgic_its_trigger_msi(kvm, iodev->its, msi->devid, msi->data);
+ mutex_unlock(&iodev->its->its_lock);
+
+ return 0;
+}
+
/* Requires the its_lock to be held. */
static void its_free_itte(struct kvm *kvm, struct its_itte *itte)
{
@@ -897,6 +956,21 @@ static int vgic_its_cmd_handle_movall(struct kvm *kvm, struct vgic_its *its,
}
/*
+ * The INT command injects the LPI associated with that DevID/EvID pair.
+ * Must be called with the its_lock mutex held.
+ */
+static int vgic_its_cmd_handle_int(struct kvm *kvm, struct vgic_its *its,
+ u64 *its_cmd)
+{
+ u32 msi_data = its_cmd_get_id(its_cmd);
+ u64 msi_devid = its_cmd_get_deviceid(its_cmd);
+
+ vgic_its_trigger_msi(kvm, its, msi_devid, msi_data);
+
+ return 0;
+}
+
+/*
* This function is called with the its_cmd lock held, but the ITS data
* structure lock dropped.
*/
@@ -932,6 +1006,9 @@ static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
case GITS_CMD_MOVALL:
ret = vgic_its_cmd_handle_movall(kvm, its, its_cmd);
break;
+ case GITS_CMD_INT:
+ ret = vgic_its_cmd_handle_int(kvm, its, its_cmd);
+ break;
case GITS_CMD_INV:
ret = vgic_its_cmd_handle_inv(kvm, its, its_cmd);
break;
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index ee348de..9d557f2 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -78,6 +78,7 @@ int vgic_v3_map_resources(struct kvm *kvm);
int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t dist_base_address);
bool vgic_has_its(struct kvm *kvm);
void vgic_enable_lpis(struct kvm_vcpu *vcpu);
+int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi);
#else
static inline void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu)
{
@@ -138,6 +139,11 @@ static inline bool vgic_has_its(struct kvm *kvm)
static inline void vgic_enable_lpis(struct kvm_vcpu *vcpu)
{
}
+
+static inline int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
+{
+ return -ENODEV;
+}
#endif
int kvm_register_vgic_device(unsigned long type);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation
2016-07-22 17:28 ` [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation Marc Zyngier
@ 2016-08-01 18:20 ` Christoffer Dall
2016-08-02 10:18 ` Marc Zyngier
2016-08-04 10:47 ` Christoffer Dall
1 sibling, 1 reply; 67+ messages in thread
From: Christoffer Dall @ 2016-08-01 18:20 UTC (permalink / raw)
To: linux-arm-kernel
On Fri, Jul 22, 2016 at 06:28:58PM +0100, Marc Zyngier wrote:
> From: Andre Przywara <andre.przywara@arm.com>
>
> When userland wants to inject an MSI into the guest, it uses the
> KVM_SIGNAL_MSI ioctl, which carries the doorbell address along with
> the payload and the device ID.
> With the help of the KVM IO bus framework we learn the corresponding
> ITS from the doorbell address. We then use our wrapper functions to
> iterate the linked lists and find the proper Interrupt Translation Table
> Entry (ITTE) and thus the corresponding struct vgic_irq to finally set
> the pending bit.
> We also provide the handler for the ITS "INT" command, which allows a
> guest to trigger an MSI via the ITS command queue. Since this one knows
> about the right ITS already, we directly call the MMIO handler function
> without using the kvm_io_bus framework.
>
> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
> Tested-by: Eric Auger <eric.auger@redhat.com>
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> ---
> virt/kvm/arm/vgic/vgic-its.c | 77 ++++++++++++++++++++++++++++++++++++++++++++
> virt/kvm/arm/vgic/vgic.h | 6 ++++
> 2 files changed, 83 insertions(+)
>
> diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
> index 1408c88..d8e8f14 100644
> --- a/virt/kvm/arm/vgic/vgic-its.c
> +++ b/virt/kvm/arm/vgic/vgic-its.c
> @@ -437,6 +437,65 @@ static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
> return 0;
> }
>
> +/*
> + * Find the target VCPU and the LPI number for a given devid/eventid pair
> + * and make this IRQ pending, possibly injecting it.
> + * Must be called with the its_lock mutex held.
> + */
> +static void vgic_its_trigger_msi(struct kvm *kvm, struct vgic_its *its,
> + u32 devid, u32 eventid)
> +{
> + struct its_itte *itte;
> +
> + if (!its->enabled)
> + return;
> +
> + itte = find_itte(its, devid, eventid);
> + /* Triggering an unmapped IRQ gets silently dropped. */
> + if (itte && its_is_collection_mapped(itte->collection)) {
> + struct kvm_vcpu *vcpu;
> +
> + vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
> + if (vcpu && vcpu->arch.vgic_cpu.lpis_enabled) {
> + spin_lock(&itte->irq->irq_lock);
> + itte->irq->pending = true;
> + vgic_queue_irq_unlock(kvm, itte->irq);
> + }
> + }
> +}
> +
> +/*
> + * Queries the KVM IO bus framework to get the ITS pointer from the given
> + * doorbell address.
> + * We then call vgic_its_trigger_msi() with the decoded data.
> + */
> +int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
> +{
> + u64 address;
> + struct kvm_io_device *kvm_io_dev;
> + struct vgic_io_device *iodev;
> +
> + if (!vgic_has_its(kvm))
> + return -ENODEV;
> +
> + if (!(msi->flags & KVM_MSI_VALID_DEVID))
> + return -EINVAL;
> +
> + address = (u64)msi->address_hi << 32 | msi->address_lo;
> +
> + kvm_io_dev = kvm_io_bus_get_dev(kvm, KVM_MMIO_BUS, address);
> + if (!kvm_io_dev)
> + return -ENODEV;
> +
> + iodev = container_of(kvm_io_dev, struct vgic_io_device, dev);
> +
> + mutex_lock(&iodev->its->its_lock);
> + vgic_its_trigger_msi(kvm, iodev->its, msi->devid, msi->data);
> + mutex_unlock(&iodev->its->its_lock);
> +
> + return 0;
According to KVM_SIGNAL_MSI this means that the guest blocked the MSI.
Is this the intention, or is the return value translated somewhere
higher in the call stack?
> +}
> +
> /* Requires the its_lock to be held. */
> static void its_free_itte(struct kvm *kvm, struct its_itte *itte)
> {
> @@ -897,6 +956,21 @@ static int vgic_its_cmd_handle_movall(struct kvm *kvm, struct vgic_its *its,
> }
>
> /*
> + * The INT command injects the LPI associated with that DevID/EvID pair.
> + * Must be called with the its_lock mutex held.
> + */
> +static int vgic_its_cmd_handle_int(struct kvm *kvm, struct vgic_its *its,
> + u64 *its_cmd)
> +{
> + u32 msi_data = its_cmd_get_id(its_cmd);
> + u64 msi_devid = its_cmd_get_deviceid(its_cmd);
> +
> + vgic_its_trigger_msi(kvm, its, msi_devid, msi_data);
> +
> + return 0;
> +}
> +
> +/*
> * This function is called with the its_cmd lock held, but the ITS data
> * structure lock dropped.
> */
> @@ -932,6 +1006,9 @@ static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
> case GITS_CMD_MOVALL:
> ret = vgic_its_cmd_handle_movall(kvm, its, its_cmd);
> break;
> + case GITS_CMD_INT:
> + ret = vgic_its_cmd_handle_int(kvm, its, its_cmd);
> + break;
> case GITS_CMD_INV:
> ret = vgic_its_cmd_handle_inv(kvm, its, its_cmd);
> break;
> diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
> index ee348de..9d557f2 100644
> --- a/virt/kvm/arm/vgic/vgic.h
> +++ b/virt/kvm/arm/vgic/vgic.h
> @@ -78,6 +78,7 @@ int vgic_v3_map_resources(struct kvm *kvm);
> int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t dist_base_address);
> bool vgic_has_its(struct kvm *kvm);
> void vgic_enable_lpis(struct kvm_vcpu *vcpu);
> +int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi);
> #else
> static inline void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu)
> {
> @@ -138,6 +139,11 @@ static inline bool vgic_has_its(struct kvm *kvm)
> static inline void vgic_enable_lpis(struct kvm_vcpu *vcpu)
> {
> }
> +
> +static inline int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
> +{
> + return -ENODEV;
> +}
> #endif
>
> int kvm_register_vgic_device(unsigned long type);
> --
> 2.8.1
>
^ permalink raw reply [flat|nested] 67+ messages in thread* [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation
2016-08-01 18:20 ` Christoffer Dall
@ 2016-08-02 10:18 ` Marc Zyngier
0 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-08-02 10:18 UTC (permalink / raw)
To: linux-arm-kernel
On 01/08/16 19:20, Christoffer Dall wrote:
> On Fri, Jul 22, 2016 at 06:28:58PM +0100, Marc Zyngier wrote:
>> From: Andre Przywara <andre.przywara@arm.com>
>>
>> When userland wants to inject an MSI into the guest, it uses the
>> KVM_SIGNAL_MSI ioctl, which carries the doorbell address along with
>> the payload and the device ID.
>> With the help of the KVM IO bus framework we learn the corresponding
>> ITS from the doorbell address. We then use our wrapper functions to
>> iterate the linked lists and find the proper Interrupt Translation Table
>> Entry (ITTE) and thus the corresponding struct vgic_irq to finally set
>> the pending bit.
>> We also provide the handler for the ITS "INT" command, which allows a
>> guest to trigger an MSI via the ITS command queue. Since this one knows
>> about the right ITS already, we directly call the MMIO handler function
>> without using the kvm_io_bus framework.
>>
>> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
>> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
>> Tested-by: Eric Auger <eric.auger@redhat.com>
>> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
>> ---
>> virt/kvm/arm/vgic/vgic-its.c | 77 ++++++++++++++++++++++++++++++++++++++++++++
>> virt/kvm/arm/vgic/vgic.h | 6 ++++
>> 2 files changed, 83 insertions(+)
>>
>> diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
>> index 1408c88..d8e8f14 100644
>> --- a/virt/kvm/arm/vgic/vgic-its.c
>> +++ b/virt/kvm/arm/vgic/vgic-its.c
>> @@ -437,6 +437,65 @@ static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
>> return 0;
>> }
>>
>> +/*
>> + * Find the target VCPU and the LPI number for a given devid/eventid pair
>> + * and make this IRQ pending, possibly injecting it.
>> + * Must be called with the its_lock mutex held.
>> + */
>> +static void vgic_its_trigger_msi(struct kvm *kvm, struct vgic_its *its,
>> + u32 devid, u32 eventid)
>> +{
>> + struct its_itte *itte;
>> +
>> + if (!its->enabled)
>> + return;
>> +
>> + itte = find_itte(its, devid, eventid);
>> + /* Triggering an unmapped IRQ gets silently dropped. */
>> + if (itte && its_is_collection_mapped(itte->collection)) {
>> + struct kvm_vcpu *vcpu;
>> +
>> + vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
>> + if (vcpu && vcpu->arch.vgic_cpu.lpis_enabled) {
>> + spin_lock(&itte->irq->irq_lock);
>> + itte->irq->pending = true;
>> + vgic_queue_irq_unlock(kvm, itte->irq);
>> + }
>> + }
>> +}
>> +
>> +/*
>> + * Queries the KVM IO bus framework to get the ITS pointer from the given
>> + * doorbell address.
>> + * We then call vgic_its_trigger_msi() with the decoded data.
>> + */
>> +int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
>> +{
>> + u64 address;
>> + struct kvm_io_device *kvm_io_dev;
>> + struct vgic_io_device *iodev;
>> +
>> + if (!vgic_has_its(kvm))
>> + return -ENODEV;
>> +
>> + if (!(msi->flags & KVM_MSI_VALID_DEVID))
>> + return -EINVAL;
>> +
>> + address = (u64)msi->address_hi << 32 | msi->address_lo;
>> +
>> + kvm_io_dev = kvm_io_bus_get_dev(kvm, KVM_MMIO_BUS, address);
>> + if (!kvm_io_dev)
>> + return -ENODEV;
>> +
>> + iodev = container_of(kvm_io_dev, struct vgic_io_device, dev);
>> +
>> + mutex_lock(&iodev->its->its_lock);
>> + vgic_its_trigger_msi(kvm, iodev->its, msi->devid, msi->data);
>> + mutex_unlock(&iodev->its->its_lock);
>> +
>> + return 0;
>
> According to KVM_SIGNAL_MSI this means that the guest blocked the MSI.
> Is this the intention, or is the return value translated somewhere
> higher in the call stack?
I'm afraid it is not, and this is definitely a bug. Andre, can you
please address this one urgently, as this has a direct impact on the
userspace API?
Thanks,
M.
--
Jazz is not dead. It just smells funny...
^ permalink raw reply [flat|nested] 67+ messages in thread
* [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation
2016-07-22 17:28 ` [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation Marc Zyngier
2016-08-01 18:20 ` Christoffer Dall
@ 2016-08-04 10:47 ` Christoffer Dall
1 sibling, 0 replies; 67+ messages in thread
From: Christoffer Dall @ 2016-08-04 10:47 UTC (permalink / raw)
To: linux-arm-kernel
Hi Andre and Marc,
On Fri, Jul 22, 2016 at 06:28:58PM +0100, Marc Zyngier wrote:
> From: Andre Przywara <andre.przywara@arm.com>
>
> When userland wants to inject an MSI into the guest, it uses the
> KVM_SIGNAL_MSI ioctl, which carries the doorbell address along with
> the payload and the device ID.
> With the help of the KVM IO bus framework we learn the corresponding
> ITS from the doorbell address. We then use our wrapper functions to
> iterate the linked lists and find the proper Interrupt Translation Table
> Entry (ITTE) and thus the corresponding struct vgic_irq to finally set
> the pending bit.
> We also provide the handler for the ITS "INT" command, which allows a
> guest to trigger an MSI via the ITS command queue. Since this one knows
> about the right ITS already, we directly call the MMIO handler function
> without using the kvm_io_bus framework.
>
> Signed-off-by: Andre Przywara <andre.przywara@arm.com>
> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
> Tested-by: Eric Auger <eric.auger@redhat.com>
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> ---
> virt/kvm/arm/vgic/vgic-its.c | 77 ++++++++++++++++++++++++++++++++++++++++++++
> virt/kvm/arm/vgic/vgic.h | 6 ++++
> 2 files changed, 83 insertions(+)
>
> diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
> index 1408c88..d8e8f14 100644
> --- a/virt/kvm/arm/vgic/vgic-its.c
> +++ b/virt/kvm/arm/vgic/vgic-its.c
> @@ -437,6 +437,65 @@ static unsigned long vgic_mmio_read_its_idregs(struct kvm *kvm,
> return 0;
> }
>
> +/*
> + * Find the target VCPU and the LPI number for a given devid/eventid pair
> + * and make this IRQ pending, possibly injecting it.
> + * Must be called with the its_lock mutex held.
> + */
> +static void vgic_its_trigger_msi(struct kvm *kvm, struct vgic_its *its,
> + u32 devid, u32 eventid)
> +{
> + struct its_itte *itte;
> +
> + if (!its->enabled)
> + return;
> +
> + itte = find_itte(its, devid, eventid);
> + /* Triggering an unmapped IRQ gets silently dropped. */
> + if (itte && its_is_collection_mapped(itte->collection)) {
> + struct kvm_vcpu *vcpu;
> +
> + vcpu = kvm_get_vcpu(kvm, itte->collection->target_addr);
> + if (vcpu && vcpu->arch.vgic_cpu.lpis_enabled) {
> + spin_lock(&itte->irq->irq_lock);
> + itte->irq->pending = true;
> + vgic_queue_irq_unlock(kvm, itte->irq);
> + }
> + }
> +}
> +
> +/*
> + * Queries the KVM IO bus framework to get the ITS pointer from the given
> + * doorbell address.
> + * We then call vgic_its_trigger_msi() with the decoded data.
> + */
> +int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi)
> +{
> + u64 address;
> + struct kvm_io_device *kvm_io_dev;
> + struct vgic_io_device *iodev;
> +
> + if (!vgic_has_its(kvm))
> + return -ENODEV;
> +
> + if (!(msi->flags & KVM_MSI_VALID_DEVID))
> + return -EINVAL;
> +
> + address = (u64)msi->address_hi << 32 | msi->address_lo;
> +
> + kvm_io_dev = kvm_io_bus_get_dev(kvm, KVM_MMIO_BUS, address);
> + if (!kvm_io_dev)
> + return -ENODEV;
> +
> + iodev = container_of(kvm_io_dev, struct vgic_io_device, dev);
This container_of bothers me; as far as I can tell, the address we use
to look stuff up on the KVM_MMIO_BUS is purely given by userspace and we
can have other stuff registered on the KVM_MMIO_BUS, which is not
contained in a struct vgic_io_device (eventfd for example).
This smells like something that could be exploited, so if I'm right, we
should plug this for -rc2 in some way.
Thanks,
-Christoffer
^ permalink raw reply [flat|nested] 67+ messages in thread
* [PATCH 42/55] KVM: arm64: vgic-its: Enable ITS emulation as a virtual MSI controller
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (40 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 41/55] KVM: arm64: vgic-its: Implement MSI injection in ITS emulation Marc Zyngier
@ 2016-07-22 17:28 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 43/55] KVM: arm/arm64: Fix vGICv2 KVM_DEV_ARM_VGIC_GRP_CPU/DIST_REGS Marc Zyngier
` (13 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:28 UTC (permalink / raw)
To: linux-arm-kernel
From: Andre Przywara <andre.przywara@arm.com>
Now that all ITS emulation functionality is in place, we advertise
MSI functionality to userland and also the ITS device to the guest - if
userland has configured that.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
Documentation/virtual/kvm/api.txt | 2 +-
arch/arm64/kvm/Kconfig | 1 +
arch/arm64/kvm/Makefile | 1 +
arch/arm64/kvm/reset.c | 6 ++++++
include/kvm/arm_vgic.h | 5 +++++
virt/kvm/arm/vgic/vgic-init.c | 3 +++
virt/kvm/arm/vgic/vgic-kvm-device.c | 3 +++
virt/kvm/arm/vgic/vgic-mmio-v3.c | 14 ++++++++++----
virt/kvm/arm/vgic/vgic.c | 8 ++++++++
virt/kvm/arm/vgic/vgic.h | 6 ++++++
10 files changed, 44 insertions(+), 5 deletions(-)
diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index 6551311..07049ea 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -2162,7 +2162,7 @@ after pausing the vcpu, but before it is resumed.
4.71 KVM_SIGNAL_MSI
Capability: KVM_CAP_SIGNAL_MSI
-Architectures: x86
+Architectures: x86 arm64
Type: vm ioctl
Parameters: struct kvm_msi (in)
Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
index aa2e34e..9d2eff0 100644
--- a/arch/arm64/kvm/Kconfig
+++ b/arch/arm64/kvm/Kconfig
@@ -36,6 +36,7 @@ config KVM
select HAVE_KVM_IRQFD
select KVM_ARM_VGIC_V3
select KVM_ARM_PMU if HW_PERF_EVENTS
+ select HAVE_KVM_MSI
---help---
Support hosting virtualized guest machines.
We don't support KVM with 16K page tables yet, due to the multiple
diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
index f00b2cd..a5b9664 100644
--- a/arch/arm64/kvm/Makefile
+++ b/arch/arm64/kvm/Makefile
@@ -29,5 +29,6 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v2.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v3.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-kvm-device.o
+kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
index e95d4f6..5bc4608 100644
--- a/arch/arm64/kvm/reset.c
+++ b/arch/arm64/kvm/reset.c
@@ -86,6 +86,12 @@ int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
case KVM_CAP_VCPU_ATTRIBUTES:
r = 1;
break;
+ case KVM_CAP_MSI_DEVID:
+ if (!kvm)
+ r = -EINVAL;
+ else
+ r = kvm->arch.vgic.msis_require_devid;
+ break;
default:
r = 0;
}
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index a6ca326..4e63a07 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -163,6 +163,9 @@ struct vgic_dist {
/* vGIC model the kernel emulates for the guest (GICv2 or GICv3) */
u32 vgic_model;
+ /* Do injected MSIs require an additional device ID? */
+ bool msis_require_devid;
+
int nr_spis;
/* TODO: Consider moving to global state */
@@ -308,4 +311,6 @@ static inline int kvm_vgic_get_max_vcpus(void)
return kvm_vgic_global_state.max_gic_vcpus;
}
+int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);
+
#endif /* __KVM_ARM_VGIC_H */
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 535e713..01a60dc 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -258,6 +258,9 @@ int vgic_init(struct kvm *kvm)
if (ret)
goto out;
+ if (vgic_has_its(kvm))
+ dist->msis_require_devid = true;
+
kvm_for_each_vcpu(i, vcpu, kvm)
kvm_vgic_vcpu_init(vcpu);
diff --git a/virt/kvm/arm/vgic/vgic-kvm-device.c b/virt/kvm/arm/vgic/vgic-kvm-device.c
index 561d2ba..1813f93 100644
--- a/virt/kvm/arm/vgic/vgic-kvm-device.c
+++ b/virt/kvm/arm/vgic/vgic-kvm-device.c
@@ -223,6 +223,9 @@ int kvm_register_vgic_device(unsigned long type)
case KVM_DEV_TYPE_ARM_VGIC_V3:
ret = kvm_register_device_ops(&kvm_arm_vgic_v3_ops,
KVM_DEV_TYPE_ARM_VGIC_V3);
+ if (ret)
+ break;
+ ret = kvm_vgic_register_its_device();
break;
#endif
}
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c
index 84a301d..ff668e0 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v3.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c
@@ -66,7 +66,12 @@ static unsigned long vgic_mmio_read_v3_misc(struct kvm_vcpu *vcpu,
case GICD_TYPER:
value = vcpu->kvm->arch.vgic.nr_spis + VGIC_NR_PRIVATE_IRQS;
value = (value >> 5) - 1;
- value |= (INTERRUPT_ID_BITS_SPIS - 1) << 19;
+ if (vgic_has_its(vcpu->kvm)) {
+ value |= (INTERRUPT_ID_BITS_ITS - 1) << 19;
+ value |= GICD_TYPER_LPIS;
+ } else {
+ value |= (INTERRUPT_ID_BITS_SPIS - 1) << 19;
+ }
break;
case GICD_IIDR:
value = (PRODUCT_ID_KVM << 24) | (IMPLEMENTER_ARM << 0);
@@ -163,9 +168,8 @@ static void vgic_mmio_write_v3r_ctlr(struct kvm_vcpu *vcpu,
vgic_cpu->lpis_enabled = val & GICR_CTLR_ENABLE_LPIS;
- if (!was_enabled && vgic_cpu->lpis_enabled) {
- /* Eventually do something */
- }
+ if (!was_enabled && vgic_cpu->lpis_enabled)
+ vgic_enable_lpis(vcpu);
}
static unsigned long vgic_mmio_read_v3r_typer(struct kvm_vcpu *vcpu,
@@ -179,6 +183,8 @@ static unsigned long vgic_mmio_read_v3r_typer(struct kvm_vcpu *vcpu,
value |= ((target_vcpu_id & 0xffff) << 8);
if (target_vcpu_id == atomic_read(&vcpu->kvm->online_vcpus) - 1)
value |= GICR_TYPER_LAST;
+ if (vgic_has_its(vcpu->kvm))
+ value |= GICR_TYPER_PLPIS;
return extract_bytes(value, addr & 7, len);
}
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index 53299fc..424cb9c 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -718,3 +718,11 @@ bool kvm_vgic_map_is_active(struct kvm_vcpu *vcpu, unsigned int virt_irq)
return map_is_active;
}
+
+int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi)
+{
+ if (vgic_has_its(kvm))
+ return vgic_its_inject_msi(kvm, msi);
+ else
+ return -ENODEV;
+}
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 9d557f2..9d40d7b 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -77,6 +77,7 @@ int vgic_v3_probe(const struct gic_kvm_info *info);
int vgic_v3_map_resources(struct kvm *kvm);
int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t dist_base_address);
bool vgic_has_its(struct kvm *kvm);
+int kvm_vgic_register_its_device(void);
void vgic_enable_lpis(struct kvm_vcpu *vcpu);
int vgic_its_inject_msi(struct kvm *kvm, struct kvm_msi *msi);
#else
@@ -136,6 +137,11 @@ static inline bool vgic_has_its(struct kvm *kvm)
return false;
}
+static inline int kvm_vgic_register_its_device(void)
+{
+ return -ENODEV;
+}
+
static inline void vgic_enable_lpis(struct kvm_vcpu *vcpu)
{
}
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 43/55] KVM: arm/arm64: Fix vGICv2 KVM_DEV_ARM_VGIC_GRP_CPU/DIST_REGS
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (41 preceding siblings ...)
2016-07-22 17:28 ` [PATCH 42/55] KVM: arm64: vgic-its: Enable ITS emulation as a virtual MSI controller Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 44/55] irqchip/gicv3-its: Restore all cacheability attributes Marc Zyngier
` (12 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
From: Eric Auger <eric.auger@redhat.com>
For VGICv2 save and restore the CPU interface registers
are accessed. Restore the modality which has been altered.
Also explicitly set the iodev_type for both the DIST and CPU
interface.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-mmio-v2.c | 2 ++
virt/kvm/arm/vgic/vgic-mmio.c | 4 +++-
2 files changed, 5 insertions(+), 1 deletion(-)
diff --git a/virt/kvm/arm/vgic/vgic-mmio-v2.c b/virt/kvm/arm/vgic/vgic-mmio-v2.c
index 4152348..b44b359 100644
--- a/virt/kvm/arm/vgic/vgic-mmio-v2.c
+++ b/virt/kvm/arm/vgic/vgic-mmio-v2.c
@@ -437,6 +437,7 @@ int vgic_v2_cpuif_uaccess(struct kvm_vcpu *vcpu, bool is_write,
struct vgic_io_device dev = {
.regions = vgic_v2_cpu_registers,
.nr_regions = ARRAY_SIZE(vgic_v2_cpu_registers),
+ .iodev_type = IODEV_CPUIF,
};
return vgic_uaccess(vcpu, &dev, is_write, offset, val);
@@ -448,6 +449,7 @@ int vgic_v2_dist_uaccess(struct kvm_vcpu *vcpu, bool is_write,
struct vgic_io_device dev = {
.regions = vgic_v2_dist_registers,
.nr_regions = ARRAY_SIZE(vgic_v2_dist_registers),
+ .iodev_type = IODEV_DIST,
};
return vgic_uaccess(vcpu, &dev, is_write, offset, val);
diff --git a/virt/kvm/arm/vgic/vgic-mmio.c b/virt/kvm/arm/vgic/vgic-mmio.c
index 26be827..3bad3c5 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.c
+++ b/virt/kvm/arm/vgic/vgic-mmio.c
@@ -484,7 +484,8 @@ static int dispatch_mmio_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
switch (iodev->iodev_type) {
case IODEV_CPUIF:
- return 1;
+ data = region->read(vcpu, addr, len);
+ break;
case IODEV_DIST:
data = region->read(vcpu, addr, len);
break;
@@ -517,6 +518,7 @@ static int dispatch_mmio_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
switch (iodev->iodev_type) {
case IODEV_CPUIF:
+ region->write(vcpu, addr, len, data);
break;
case IODEV_DIST:
region->write(vcpu, addr, len, data);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 44/55] irqchip/gicv3-its: Restore all cacheability attributes
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (42 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 43/55] KVM: arm/arm64: Fix vGICv2 KVM_DEV_ARM_VGIC_GRP_CPU/DIST_REGS Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 45/55] KVM: arm64: vgic-its: Generalize use of vgic_get_irq_kref Marc Zyngier
` (11 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
Let's restore some of the #defines that have been savagely dropped
by the introduction of the KVM ITS code, as pointlessly break
other users (including series that are already in -next).
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/linux/irqchip/arm-gic-v3.h | 48 +++++++++++++++++++++++++++++++-------
1 file changed, 40 insertions(+), 8 deletions(-)
diff --git a/include/linux/irqchip/arm-gic-v3.h b/include/linux/irqchip/arm-gic-v3.h
index 9442be7..700b421 100644
--- a/include/linux/irqchip/arm-gic-v3.h
+++ b/include/linux/irqchip/arm-gic-v3.h
@@ -146,8 +146,16 @@
#define GICR_PROPBASER_InnerShareable \
GIC_BASER_SHAREABILITY(GICR_PROPBASER, InnerShareable)
-#define GICR_PROPBASER_nC GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, nC)
-#define GICR_PROPBASER_WaWb GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, WaWb)
+
+#define GICR_PROPBASER_nCnB GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, nCnB)
+#define GICR_PROPBASER_nC GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, nC)
+#define GICR_PROPBASER_RaWt GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWt)
+#define GICR_PROPBASER_RaWb GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWt)
+#define GICR_PROPBASER_WaWt GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, WaWt)
+#define GICR_PROPBASER_WaWb GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, WaWb)
+#define GICR_PROPBASER_RaWaWt GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWaWt)
+#define GICR_PROPBASER_RaWaWb GIC_BASER_CACHEABILITY(GICR_PROPBASER, INNER, RaWaWb)
+
#define GICR_PROPBASER_IDBITS_MASK (0x1f)
#define GICR_PENDBASER_SHAREABILITY_SHIFT (10)
@@ -163,8 +171,16 @@
#define GICR_PENDBASER_InnerShareable \
GIC_BASER_SHAREABILITY(GICR_PENDBASER, InnerShareable)
-#define GICR_PENDBASER_nC GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, nC)
-#define GICR_PENDBASER_WaWb GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, WaWb)
+
+#define GICR_PENDBASER_nCnB GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, nCnB)
+#define GICR_PENDBASER_nC GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, nC)
+#define GICR_PENDBASER_RaWt GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWt)
+#define GICR_PENDBASER_RaWb GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWt)
+#define GICR_PENDBASER_WaWt GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, WaWt)
+#define GICR_PENDBASER_WaWb GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, WaWb)
+#define GICR_PENDBASER_RaWaWt GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWaWt)
+#define GICR_PENDBASER_RaWaWb GIC_BASER_CACHEABILITY(GICR_PENDBASER, INNER, RaWaWb)
+
#define GICR_PENDBASER_PTZ BIT_ULL(62)
/*
@@ -237,24 +253,40 @@
#define GITS_CBASER_InnerShareable \
GIC_BASER_SHAREABILITY(GITS_CBASER, InnerShareable)
-#define GITS_CBASER_nC GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, nC)
-#define GITS_CBASER_WaWb GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, WaWb)
+
+#define GITS_CBASER_nCnB GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, nCnB)
+#define GITS_CBASER_nC GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, nC)
+#define GITS_CBASER_RaWt GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWt)
+#define GITS_CBASER_RaWb GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWt)
+#define GITS_CBASER_WaWt GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, WaWt)
+#define GITS_CBASER_WaWb GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, WaWb)
+#define GITS_CBASER_RaWaWt GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWaWt)
+#define GITS_CBASER_RaWaWb GIC_BASER_CACHEABILITY(GITS_CBASER, INNER, RaWaWb)
#define GITS_BASER_NR_REGS 8
#define GITS_BASER_VALID (1UL << 63)
#define GITS_BASER_INDIRECT (1ULL << 62)
+
#define GITS_BASER_INNER_CACHEABILITY_SHIFT (59)
#define GITS_BASER_OUTER_CACHEABILITY_SHIFT (53)
#define GITS_BASER_INNER_CACHEABILITY_MASK \
GIC_BASER_CACHEABILITY(GITS_BASER, INNER, MASK)
+#define GITS_BASER_CACHEABILITY_MASK GITS_BASER_INNER_CACHEABILITY_MASK
#define GITS_BASER_OUTER_CACHEABILITY_MASK \
GIC_BASER_CACHEABILITY(GITS_BASER, OUTER, MASK)
#define GITS_BASER_SHAREABILITY_MASK \
GIC_BASER_SHAREABILITY(GITS_BASER, SHAREABILITY_MASK)
-#define GITS_BASER_nC GIC_BASER_CACHEABILITY(GITS_BASER, INNER, nC)
-#define GITS_BASER_WaWb GIC_BASER_CACHEABILITY(GITS_BASER, INNER, WaWb)
+#define GITS_BASER_nCnB GIC_BASER_CACHEABILITY(GITS_BASER, INNER, nCnB)
+#define GITS_BASER_nC GIC_BASER_CACHEABILITY(GITS_BASER, INNER, nC)
+#define GITS_BASER_RaWt GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWt)
+#define GITS_BASER_RaWb GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWt)
+#define GITS_BASER_WaWt GIC_BASER_CACHEABILITY(GITS_BASER, INNER, WaWt)
+#define GITS_BASER_WaWb GIC_BASER_CACHEABILITY(GITS_BASER, INNER, WaWb)
+#define GITS_BASER_RaWaWt GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWaWt)
+#define GITS_BASER_RaWaWb GIC_BASER_CACHEABILITY(GITS_BASER, INNER, RaWaWb)
+
#define GITS_BASER_TYPE_SHIFT (56)
#define GITS_BASER_TYPE(r) (((r) >> GITS_BASER_TYPE_SHIFT) & 7)
#define GITS_BASER_ENTRY_SIZE_SHIFT (48)
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 45/55] KVM: arm64: vgic-its: Generalize use of vgic_get_irq_kref
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (43 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 44/55] irqchip/gicv3-its: Restore all cacheability attributes Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 46/55] KVM: arm64: vgic-its: Fix handling of indirect tables Marc Zyngier
` (10 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
Instead of sprinkling raw kref_get() calls everytime we cannot
do a normal vgic_get_irq(), use the existing vgic_get_irq_kref(),
which does the same thing and is paired with a vgic_put_irq().
vgic_get_irq_kref is moved to vgic.h in order to be easily shared.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 2 +-
virt/kvm/arm/vgic/vgic.c | 10 +---------
virt/kvm/arm/vgic/vgic.h | 8 ++++++++
3 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index d8e8f14..f427fa2 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -80,7 +80,7 @@ static struct vgic_irq *vgic_add_lpi(struct kvm *kvm, u32 intid)
* call vgic_put_irq() on the returned pointer once it's
* finished with the IRQ.
*/
- kref_get(&irq->refcount);
+ vgic_get_irq_kref(irq);
goto out_unlock;
}
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index 424cb9c..39f3358 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -71,7 +71,7 @@ static struct vgic_irq *vgic_get_lpi(struct kvm *kvm, u32 intid)
* This increases the refcount, the caller is expected to
* call vgic_put_irq() later once it's finished with the IRQ.
*/
- kref_get(&irq->refcount);
+ vgic_get_irq_kref(irq);
goto out_unlock;
}
irq = NULL;
@@ -106,14 +106,6 @@ struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu,
return NULL;
}
-static void vgic_get_irq_kref(struct vgic_irq *irq)
-{
- if (irq->intid < VGIC_MIN_LPI)
- return;
-
- kref_get(&irq->refcount);
-}
-
/*
* We can't do anything in here, because we lack the kvm pointer to
* lock and remove the item from the lpi_list. So we keep this function
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 9d40d7b..1d8e21d 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -64,6 +64,14 @@ int vgic_v2_map_resources(struct kvm *kvm);
int vgic_register_dist_iodev(struct kvm *kvm, gpa_t dist_base_address,
enum vgic_type);
+static inline void vgic_get_irq_kref(struct vgic_irq *irq)
+{
+ if (irq->intid < VGIC_MIN_LPI)
+ return;
+
+ kref_get(&irq->refcount);
+}
+
#ifdef CONFIG_KVM_ARM_VGIC_V3
void vgic_v3_process_maintenance(struct kvm_vcpu *vcpu);
void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 46/55] KVM: arm64: vgic-its: Fix handling of indirect tables
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (44 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 45/55] KVM: arm64: vgic-its: Generalize use of vgic_get_irq_kref Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 47/55] KVM: arm64: vgic-its: Fix vgic_its_check_device_id BE handling Marc Zyngier
` (9 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
The current code will fail on valid indirect tables, and happily
use the ones that are pointing out of the guest RAM. Funny what a
small "!" can do for you...
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index f427fa2..d6697c4 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -702,9 +702,9 @@ static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
return false;
/* Each 1st level entry is represented by a 64-bit value. */
- if (!kvm_read_guest(kvm,
- BASER_ADDRESS(r) + index * sizeof(indirect_ptr),
- &indirect_ptr, sizeof(indirect_ptr)))
+ if (kvm_read_guest(kvm,
+ BASER_ADDRESS(r) + index * sizeof(indirect_ptr),
+ &indirect_ptr, sizeof(indirect_ptr)))
return false;
/* check the valid bit of the first level entry */
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 47/55] KVM: arm64: vgic-its: Fix vgic_its_check_device_id BE handling
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (45 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 46/55] KVM: arm64: vgic-its: Fix handling of indirect tables Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 48/55] KVM: arm64: vgic-its: Fix misleading nr_entries in vgic_its_check_device_id Marc Zyngier
` (8 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
The ITS tables are stored in LE format. If the host is reading
a L1 table entry to check its validity, it must convert it to
the CPU endianness.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index d6697c4..2ac5927 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -707,6 +707,8 @@ static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
&indirect_ptr, sizeof(indirect_ptr)))
return false;
+ indirect_ptr = le64_to_cpu(indirect_ptr);
+
/* check the valid bit of the first level entry */
if (!(indirect_ptr & BIT_ULL(63)))
return false;
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 48/55] KVM: arm64: vgic-its: Fix misleading nr_entries in vgic_its_check_device_id
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (46 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 47/55] KVM: arm64: vgic-its: Fix vgic_its_check_device_id BE handling Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 49/55] KVM: arm64: vgic-its: Validate the device table L1 entry Marc Zyngier
` (7 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
The nr_entries variable in vgic_its_check_device_id actually
describe the size of the L1 table, and not the number of
entries in this table.
Rename it to l1_tbl_size, so that we can now change the code
with a better understanding of what is what.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 2ac5927..268a0c7 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -687,18 +687,18 @@ static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
int device_id)
{
u64 r = its->baser_device_table;
- int nr_entries = GITS_BASER_NR_PAGES(r) * SZ_64K;
+ int l1_tbl_size = GITS_BASER_NR_PAGES(r) * SZ_64K;
int index;
u64 indirect_ptr;
gfn_t gfn;
if (!(r & GITS_BASER_INDIRECT))
- return device_id < (nr_entries / GITS_BASER_ENTRY_SIZE(r));
+ return device_id < (l1_tbl_size / GITS_BASER_ENTRY_SIZE(r));
/* calculate and check the index into the 1st level */
index = device_id / (SZ_64K / GITS_BASER_ENTRY_SIZE(r));
- if (index >= (nr_entries / sizeof(u64)))
+ if (index >= (l1_tbl_size / sizeof(u64)))
return false;
/* Each 1st level entry is represented by a 64-bit value. */
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 49/55] KVM: arm64: vgic-its: Validate the device table L1 entry
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (47 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 48/55] KVM: arm64: vgic-its: Fix misleading nr_entries in vgic_its_check_device_id Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 50/55] KVM: arm64: vgic-its: Fix L2 entry validation for indirect tables Marc Zyngier
` (6 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
Checking that the device_id fits if the table, and we must make
sure that the associated memory is also accessible.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 13 +++++++++++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 268a0c7..4943d6a 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -693,8 +693,17 @@ static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
gfn_t gfn;
- if (!(r & GITS_BASER_INDIRECT))
- return device_id < (l1_tbl_size / GITS_BASER_ENTRY_SIZE(r));
+ if (!(r & GITS_BASER_INDIRECT)) {
+ phys_addr_t addr;
+
+ if (device_id >= (l1_tbl_size / GITS_BASER_ENTRY_SIZE(r)))
+ return false;
+
+ addr = BASER_ADDRESS(r) + device_id * GITS_BASER_ENTRY_SIZE(r);
+ gfn = addr >> PAGE_SHIFT;
+
+ return kvm_is_visible_gfn(kvm, gfn);
+ }
/* calculate and check the index into the 1st level */
index = device_id / (SZ_64K / GITS_BASER_ENTRY_SIZE(r));
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 50/55] KVM: arm64: vgic-its: Fix L2 entry validation for indirect tables
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (48 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 49/55] KVM: arm64: vgic-its: Validate the device table L1 entry Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 51/55] KVM: arm64: vgic-its: Add collection allocator/destructor Marc Zyngier
` (5 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
When checking that the storage address of a device entry is valid,
it is critical to compute the actual address of the entry, rather
than relying on the beginning of the page to match a CPU page of
the same size: for example, if the guest places the table at the
last 64kB boundary of RAM, but RAM size isn't a multiple of 64kB...
Fix this by computing the actual offset of the device ID in the
L2 page, and check the corresponding GFN.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 4943d6a..2faf1f4 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -727,7 +727,12 @@ static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
* Any address beyond our supported 48 bits of PA will be caught
* by the actual check in the final step.
*/
- gfn = (indirect_ptr & GENMASK_ULL(51, 16)) >> PAGE_SHIFT;
+ indirect_ptr &= GENMASK_ULL(51, 16);
+
+ /* Find the address of the actual entry */
+ index = device_id % (SZ_64K / GITS_BASER_ENTRY_SIZE(r));
+ indirect_ptr += index * GITS_BASER_ENTRY_SIZE(r);
+ gfn = indirect_ptr >> PAGE_SHIFT;
return kvm_is_visible_gfn(kvm, gfn);
}
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 51/55] KVM: arm64: vgic-its: Add collection allocator/destructor
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (49 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 50/55] KVM: arm64: vgic-its: Fix L2 entry validation for indirect tables Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 52/55] KVM: arm64: vgic-its: Add pointer to corresponding kvm_device Marc Zyngier
` (4 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
Instead of spreading random allocations all over the place,
consolidate allocation/init/freeing of collections in a pair
of constructor/destructor.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 92 ++++++++++++++++++++++++++------------------
1 file changed, 54 insertions(+), 38 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 2faf1f4..d6f68e9 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -581,14 +581,45 @@ static int vgic_its_cmd_handle_movi(struct kvm *kvm, struct vgic_its *its,
return 0;
}
-static void vgic_its_init_collection(struct vgic_its *its,
- struct its_collection *collection,
+static int vgic_its_alloc_collection(struct vgic_its *its,
+ struct its_collection **colp,
u32 coll_id)
{
+ struct its_collection *collection;
+
+ collection = kzalloc(sizeof(*collection), GFP_KERNEL);
+
collection->collection_id = coll_id;
collection->target_addr = COLLECTION_NOT_MAPPED;
list_add_tail(&collection->coll_list, &its->collection_list);
+ *colp = collection;
+
+ return 0;
+}
+
+static void vgic_its_free_collection(struct vgic_its *its, u32 coll_id)
+{
+ struct its_collection *collection;
+ struct its_device *device;
+ struct its_itte *itte;
+
+ /*
+ * Clearing the mapping for that collection ID removes the
+ * entry from the list. If there wasn't any before, we can
+ * go home early.
+ */
+ collection = find_collection(its, coll_id);
+ if (!collection)
+ return;
+
+ for_each_lpi_its(device, itte, its)
+ if (itte->collection &&
+ itte->collection->collection_id == coll_id)
+ itte->collection = NULL;
+
+ list_del(&collection->coll_list);
+ kfree(collection);
}
/*
@@ -605,6 +636,7 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
struct its_device *device;
struct its_collection *collection, *new_coll = NULL;
int lpi_nr;
+ int ret;
device = find_its_device(its, device_id);
if (!device)
@@ -612,9 +644,10 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
collection = find_collection(its, coll_id);
if (!collection) {
- new_coll = kzalloc(sizeof(struct its_collection), GFP_KERNEL);
- if (!new_coll)
- return -ENOMEM;
+ ret = vgic_its_alloc_collection(its, &collection, coll_id);
+ if (ret)
+ return ret;
+ new_coll = collection;
}
if (subcmd == GITS_CMD_MAPTI)
@@ -623,27 +656,22 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
lpi_nr = event_id;
if (lpi_nr < GIC_LPI_OFFSET ||
lpi_nr >= max_lpis_propbaser(kvm->arch.vgic.propbaser)) {
- kfree(new_coll);
- return E_ITS_MAPTI_PHYSICALID_OOR;
+ ret = E_ITS_MAPTI_PHYSICALID_OOR;
+ goto err;
}
itte = find_itte(its, device_id, event_id);
if (!itte) {
itte = kzalloc(sizeof(struct its_itte), GFP_KERNEL);
if (!itte) {
- kfree(new_coll);
- return -ENOMEM;
+ ret = -ENOMEM;
+ goto err;
}
itte->event_id = event_id;
list_add_tail(&itte->itte_list, &device->itt_head);
}
- if (!collection) {
- collection = new_coll;
- vgic_its_init_collection(its, collection, coll_id);
- }
-
itte->collection = collection;
itte->lpi = lpi_nr;
itte->irq = vgic_add_lpi(kvm, lpi_nr);
@@ -657,6 +685,10 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
update_lpi_config(kvm, itte->irq, NULL);
return 0;
+err:
+ if (new_coll)
+ vgic_its_free_collection(its, coll_id);
+ return ret;
}
/* Requires the its_lock to be held. */
@@ -809,34 +841,18 @@ static int vgic_its_cmd_handle_mapc(struct kvm *kvm, struct vgic_its *its,
if (coll_id >= vgic_its_nr_collection_ids(its))
return E_ITS_MAPC_COLLECTION_OOR;
- collection = find_collection(its, coll_id);
-
if (!valid) {
- struct its_device *device;
- struct its_itte *itte;
- /*
- * Clearing the mapping for that collection ID removes the
- * entry from the list. If there wasn't any before, we can
- * go home early.
- */
- if (!collection)
- return 0;
-
- for_each_lpi_its(device, itte, its)
- if (itte->collection &&
- itte->collection->collection_id == coll_id)
- itte->collection = NULL;
-
- list_del(&collection->coll_list);
- kfree(collection);
+ vgic_its_free_collection(its, coll_id);
} else {
+ collection = find_collection(its, coll_id);
+
if (!collection) {
- collection = kzalloc(sizeof(struct its_collection),
- GFP_KERNEL);
- if (!collection)
- return -ENOMEM;
+ int ret;
- vgic_its_init_collection(its, collection, coll_id);
+ ret = vgic_its_alloc_collection(its, &collection,
+ coll_id);
+ if (ret)
+ return ret;
collection->target_addr = target_addr;
} else {
collection->target_addr = target_addr;
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 52/55] KVM: arm64: vgic-its: Add pointer to corresponding kvm_device
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (50 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 51/55] KVM: arm64: vgic-its: Add collection allocator/destructor Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 53/55] KVM: arm64: vgic-its: Turn device_id validation into generic ID validation Marc Zyngier
` (3 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
Going from the ITS structure to the corresponding KVM structure
would be quite handy at times. The kvm_device pointer that is
passed at create time is quite convenient for this, so let's
keep a copy of it in the vgic_its structure.
This will be put to a good use in subsequent patches.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
include/kvm/arm_vgic.h | 1 +
virt/kvm/arm/vgic/vgic-its.c | 1 +
2 files changed, 2 insertions(+)
diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 4e63a07..540da51 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -138,6 +138,7 @@ struct vgic_its {
bool enabled;
bool initialized;
struct vgic_io_device iodev;
+ struct kvm_device *dev;
/* These registers correspond to GITS_BASER{0,1} */
u64 baser_device_table;
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index d6f68e9..dcae567 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -1368,6 +1368,7 @@ static int vgic_its_create(struct kvm_device *dev, u32 type)
dev->kvm->arch.vgic.has_its = true;
its->initialized = false;
its->enabled = false;
+ its->dev = dev;
its->baser_device_table = INITIAL_BASER_VALUE |
((u64)GITS_BASER_TYPE_DEVICE << GITS_BASER_TYPE_SHIFT);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 53/55] KVM: arm64: vgic-its: Turn device_id validation into generic ID validation
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (51 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 52/55] KVM: arm64: vgic-its: Add pointer to corresponding kvm_device Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 54/55] KVM: arm64: vgic-its: Make vgic_its_cmd_handle_mapi similar to other handlers Marc Zyngier
` (2 subsequent siblings)
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
There is no need to have separate functions to validate devices
and collections, as the architecture doesn't really distinguish the
two, and they are supposed to be managed the same way.
Let's turn the DevID checker into a generic one.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 134 ++++++++++++++++++++-----------------------
1 file changed, 62 insertions(+), 72 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index dcae567..996e3e1 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -581,12 +581,73 @@ static int vgic_its_cmd_handle_movi(struct kvm *kvm, struct vgic_its *its,
return 0;
}
+/*
+ * Check whether an ID can be stored into the corresponding guest table.
+ * For a direct table this is pretty easy, but gets a bit nasty for
+ * indirect tables. We check whether the resulting guest physical address
+ * is actually valid (covered by a memslot and guest accessbible).
+ * For this we have to read the respective first level entry.
+ */
+static bool vgic_its_check_id(struct vgic_its *its, u64 baser, int id)
+{
+ int l1_tbl_size = GITS_BASER_NR_PAGES(baser) * SZ_64K;
+ int index;
+ u64 indirect_ptr;
+ gfn_t gfn;
+
+ if (!(baser & GITS_BASER_INDIRECT)) {
+ phys_addr_t addr;
+
+ if (id >= (l1_tbl_size / GITS_BASER_ENTRY_SIZE(baser)))
+ return false;
+
+ addr = BASER_ADDRESS(baser) + id * GITS_BASER_ENTRY_SIZE(baser);
+ gfn = addr >> PAGE_SHIFT;
+
+ return kvm_is_visible_gfn(its->dev->kvm, gfn);
+ }
+
+ /* calculate and check the index into the 1st level */
+ index = id / (SZ_64K / GITS_BASER_ENTRY_SIZE(baser));
+ if (index >= (l1_tbl_size / sizeof(u64)))
+ return false;
+
+ /* Each 1st level entry is represented by a 64-bit value. */
+ if (kvm_read_guest(its->dev->kvm,
+ BASER_ADDRESS(baser) + index * sizeof(indirect_ptr),
+ &indirect_ptr, sizeof(indirect_ptr)))
+ return false;
+
+ indirect_ptr = le64_to_cpu(indirect_ptr);
+
+ /* check the valid bit of the first level entry */
+ if (!(indirect_ptr & BIT_ULL(63)))
+ return false;
+
+ /*
+ * Mask the guest physical address and calculate the frame number.
+ * Any address beyond our supported 48 bits of PA will be caught
+ * by the actual check in the final step.
+ */
+ indirect_ptr &= GENMASK_ULL(51, 16);
+
+ /* Find the address of the actual entry */
+ index = id % (SZ_64K / GITS_BASER_ENTRY_SIZE(baser));
+ indirect_ptr += index * GITS_BASER_ENTRY_SIZE(baser);
+ gfn = indirect_ptr >> PAGE_SHIFT;
+
+ return kvm_is_visible_gfn(its->dev->kvm, gfn);
+}
+
static int vgic_its_alloc_collection(struct vgic_its *its,
struct its_collection **colp,
u32 coll_id)
{
struct its_collection *collection;
+ if (!vgic_its_check_id(its, its->baser_coll_table, coll_id))
+ return E_ITS_MAPC_COLLECTION_OOR;
+
collection = kzalloc(sizeof(*collection), GFP_KERNEL);
collection->collection_id = coll_id;
@@ -709,67 +770,6 @@ static void vgic_its_unmap_device(struct kvm *kvm, struct its_device *device)
}
/*
- * Check whether a device ID can be stored into the guest device tables.
- * For a direct table this is pretty easy, but gets a bit nasty for
- * indirect tables. We check whether the resulting guest physical address
- * is actually valid (covered by a memslot and guest accessbible).
- * For this we have to read the respective first level entry.
- */
-static bool vgic_its_check_device_id(struct kvm *kvm, struct vgic_its *its,
- int device_id)
-{
- u64 r = its->baser_device_table;
- int l1_tbl_size = GITS_BASER_NR_PAGES(r) * SZ_64K;
- int index;
- u64 indirect_ptr;
- gfn_t gfn;
-
-
- if (!(r & GITS_BASER_INDIRECT)) {
- phys_addr_t addr;
-
- if (device_id >= (l1_tbl_size / GITS_BASER_ENTRY_SIZE(r)))
- return false;
-
- addr = BASER_ADDRESS(r) + device_id * GITS_BASER_ENTRY_SIZE(r);
- gfn = addr >> PAGE_SHIFT;
-
- return kvm_is_visible_gfn(kvm, gfn);
- }
-
- /* calculate and check the index into the 1st level */
- index = device_id / (SZ_64K / GITS_BASER_ENTRY_SIZE(r));
- if (index >= (l1_tbl_size / sizeof(u64)))
- return false;
-
- /* Each 1st level entry is represented by a 64-bit value. */
- if (kvm_read_guest(kvm,
- BASER_ADDRESS(r) + index * sizeof(indirect_ptr),
- &indirect_ptr, sizeof(indirect_ptr)))
- return false;
-
- indirect_ptr = le64_to_cpu(indirect_ptr);
-
- /* check the valid bit of the first level entry */
- if (!(indirect_ptr & BIT_ULL(63)))
- return false;
-
- /*
- * Mask the guest physical address and calculate the frame number.
- * Any address beyond our supported 48 bits of PA will be caught
- * by the actual check in the final step.
- */
- indirect_ptr &= GENMASK_ULL(51, 16);
-
- /* Find the address of the actual entry */
- index = device_id % (SZ_64K / GITS_BASER_ENTRY_SIZE(r));
- indirect_ptr += index * GITS_BASER_ENTRY_SIZE(r);
- gfn = indirect_ptr >> PAGE_SHIFT;
-
- return kvm_is_visible_gfn(kvm, gfn);
-}
-
-/*
* MAPD maps or unmaps a device ID to Interrupt Translation Tables (ITTs).
* Must be called with the its_lock mutex held.
*/
@@ -780,7 +780,7 @@ static int vgic_its_cmd_handle_mapd(struct kvm *kvm, struct vgic_its *its,
bool valid = its_cmd_get_validbit(its_cmd);
struct its_device *device;
- if (!vgic_its_check_device_id(kvm, its, device_id))
+ if (!vgic_its_check_id(its, its->baser_device_table, device_id))
return E_ITS_MAPD_DEVICE_OOR;
device = find_its_device(its, device_id);
@@ -812,13 +812,6 @@ static int vgic_its_cmd_handle_mapd(struct kvm *kvm, struct vgic_its *its,
return 0;
}
-static int vgic_its_nr_collection_ids(struct vgic_its *its)
-{
- u64 r = its->baser_coll_table;
-
- return (GITS_BASER_NR_PAGES(r) * SZ_64K) / GITS_BASER_ENTRY_SIZE(r);
-}
-
/*
* The MAPC command maps collection IDs to redistributors.
* Must be called with the its_lock mutex held.
@@ -838,9 +831,6 @@ static int vgic_its_cmd_handle_mapc(struct kvm *kvm, struct vgic_its *its,
if (target_addr >= atomic_read(&kvm->online_vcpus))
return E_ITS_MAPC_PROCNUM_OOR;
- if (coll_id >= vgic_its_nr_collection_ids(its))
- return E_ITS_MAPC_COLLECTION_OOR;
-
if (!valid) {
vgic_its_free_collection(its, coll_id);
} else {
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 54/55] KVM: arm64: vgic-its: Make vgic_its_cmd_handle_mapi similar to other handlers
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (52 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 53/55] KVM: arm64: vgic-its: Turn device_id validation into generic ID validation Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 17:29 ` [PATCH 55/55] KVM: arm64: vgic-its: Simplify MAPI error handling Marc Zyngier
2016-07-22 19:50 ` [PULL] KVM/ARM updates for Linux 4.8, take #1 Radim Krčmář
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
vgic_its_cmd_handle_mapi has an extra "subcmd" argument, which is
already contained in the command buffer that all command handlers
obtain from the command queue. Let's drop it, as it is not that
useful.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 11 +++++------
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index 996e3e1..ec7e07b 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -688,7 +688,7 @@ static void vgic_its_free_collection(struct vgic_its *its, u32 coll_id)
* Must be called with its_lock mutex held.
*/
static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
- u64 *its_cmd, u8 subcmd)
+ u64 *its_cmd)
{
u32 device_id = its_cmd_get_deviceid(its_cmd);
u32 event_id = its_cmd_get_id(its_cmd);
@@ -711,7 +711,7 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
new_coll = collection;
}
- if (subcmd == GITS_CMD_MAPTI)
+ if (its_cmd_get_command(its_cmd) == GITS_CMD_MAPTI)
lpi_nr = its_cmd_get_physical_id(its_cmd);
else
lpi_nr = event_id;
@@ -999,11 +999,10 @@ static int vgic_its_cmd_handle_int(struct kvm *kvm, struct vgic_its *its,
static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
u64 *its_cmd)
{
- u8 cmd = its_cmd_get_command(its_cmd);
int ret = -ENODEV;
mutex_lock(&its->its_lock);
- switch (cmd) {
+ switch (its_cmd_get_command(its_cmd)) {
case GITS_CMD_MAPD:
ret = vgic_its_cmd_handle_mapd(kvm, its, its_cmd);
break;
@@ -1011,10 +1010,10 @@ static int vgic_its_handle_command(struct kvm *kvm, struct vgic_its *its,
ret = vgic_its_cmd_handle_mapc(kvm, its, its_cmd);
break;
case GITS_CMD_MAPI:
- ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd, cmd);
+ ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd);
break;
case GITS_CMD_MAPTI:
- ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd, cmd);
+ ret = vgic_its_cmd_handle_mapi(kvm, its, its_cmd);
break;
case GITS_CMD_MOVI:
ret = vgic_its_cmd_handle_movi(kvm, its, its_cmd);
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PATCH 55/55] KVM: arm64: vgic-its: Simplify MAPI error handling
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (53 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 54/55] KVM: arm64: vgic-its: Make vgic_its_cmd_handle_mapi similar to other handlers Marc Zyngier
@ 2016-07-22 17:29 ` Marc Zyngier
2016-07-22 19:50 ` [PULL] KVM/ARM updates for Linux 4.8, take #1 Radim Krčmář
55 siblings, 0 replies; 67+ messages in thread
From: Marc Zyngier @ 2016-07-22 17:29 UTC (permalink / raw)
To: linux-arm-kernel
If we care to move all the checks that do not involve any memory
allocation, we can simplify the MAPI error handling. Let's do that,
it cannot hurt.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic-its.c | 30 ++++++++++++------------------
1 file changed, 12 insertions(+), 18 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic-its.c b/virt/kvm/arm/vgic/vgic-its.c
index ec7e07b..07411cf 100644
--- a/virt/kvm/arm/vgic/vgic-its.c
+++ b/virt/kvm/arm/vgic/vgic-its.c
@@ -697,36 +697,34 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
struct its_device *device;
struct its_collection *collection, *new_coll = NULL;
int lpi_nr;
- int ret;
device = find_its_device(its, device_id);
if (!device)
return E_ITS_MAPTI_UNMAPPED_DEVICE;
- collection = find_collection(its, coll_id);
- if (!collection) {
- ret = vgic_its_alloc_collection(its, &collection, coll_id);
- if (ret)
- return ret;
- new_coll = collection;
- }
-
if (its_cmd_get_command(its_cmd) == GITS_CMD_MAPTI)
lpi_nr = its_cmd_get_physical_id(its_cmd);
else
lpi_nr = event_id;
if (lpi_nr < GIC_LPI_OFFSET ||
- lpi_nr >= max_lpis_propbaser(kvm->arch.vgic.propbaser)) {
- ret = E_ITS_MAPTI_PHYSICALID_OOR;
- goto err;
+ lpi_nr >= max_lpis_propbaser(kvm->arch.vgic.propbaser))
+ return E_ITS_MAPTI_PHYSICALID_OOR;
+
+ collection = find_collection(its, coll_id);
+ if (!collection) {
+ int ret = vgic_its_alloc_collection(its, &collection, coll_id);
+ if (ret)
+ return ret;
+ new_coll = collection;
}
itte = find_itte(its, device_id, event_id);
if (!itte) {
itte = kzalloc(sizeof(struct its_itte), GFP_KERNEL);
if (!itte) {
- ret = -ENOMEM;
- goto err;
+ if (new_coll)
+ vgic_its_free_collection(its, coll_id);
+ return -ENOMEM;
}
itte->event_id = event_id;
@@ -746,10 +744,6 @@ static int vgic_its_cmd_handle_mapi(struct kvm *kvm, struct vgic_its *its,
update_lpi_config(kvm, itte->irq, NULL);
return 0;
-err:
- if (new_coll)
- vgic_its_free_collection(its, coll_id);
- return ret;
}
/* Requires the its_lock to be held. */
--
2.8.1
^ permalink raw reply related [flat|nested] 67+ messages in thread* [PULL] KVM/ARM updates for Linux 4.8, take #1
2016-07-22 17:28 [PULL] KVM/ARM updates for Linux 4.8, take #1 Marc Zyngier
` (54 preceding siblings ...)
2016-07-22 17:29 ` [PATCH 55/55] KVM: arm64: vgic-its: Simplify MAPI error handling Marc Zyngier
@ 2016-07-22 19:50 ` Radim Krčmář
55 siblings, 0 replies; 67+ messages in thread
From: Radim Krčmář @ 2016-07-22 19:50 UTC (permalink / raw)
To: linux-arm-kernel
2016-07-22 18:28+0100, Marc Zyngier:
> Paolo, Radim,
>
> Please find below the first batch of 4.8 updates for KVM/ARM. Biggest
> feature is the long awaited GICv3 ITS emulation, allowing MSIs to be
> delivered into guests running on GICv3 HW. The other big feature is
> the removal of the old vgic implementation. Less visible is the revamp
> of the way we deal with the HYP initialization (by keeping an idmap
> around), and making HYP page tables honor the kernel's own protection.
>
> This may generate a few conflicts, all of which have been seen in
> -next. 3 which are KVM specific (and pretty easy to resolve), and
> another one in include/irqchip/arm-gic-v3.h. They should be resolved
> as in -next.
>
> Please pull!
Pulled into kvm/next, KVM_CAP_MSI_DEVID is 131.
> 54 files changed, 2698 insertions(+), 6065 deletions(-)
<3
^ permalink raw reply [flat|nested] 67+ messages in thread