* [RFC/PATCH v3] arm64: define MODULES_VADDR by module_alloc_base
@ 2017-08-30 11:46 Miles Chen
2017-10-17 14:43 ` Will Deacon
0 siblings, 1 reply; 3+ messages in thread
From: Miles Chen @ 2017-08-30 11:46 UTC (permalink / raw)
To: linux-arm-kernel
After the kernel ASLR, the module virtual address is moved to
[module_alloc_base, module_alloc_base + MODULES_VSIZE).
However, the MODULES_VADDR is still defined as a constant and functions
like is_vmalloc_or_module_addr() or dump functions will not able to
use correct module range information.
The patch omits modules information in virtual kenrel memory layout if
the modules area is fully mapped whitin vmalloc area.
Use module_alloc_base to define MODULES_VADDR. I tested the patch under
three different conditions:
1.CONFIG_RANDOMIZE_BASE=y, seed=0
CONFIG_KASAN=n
2.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
CONFIG_KASAN=n
3.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
CONFIG_KASAN=y
test log:
1.CONFIG_RANDOMIZE_BASE=y, seed=0
CONFIG_KASAN=n
[ 0.000000] Virtual kernel memory layout:
[ 0.000000] modules : 0xffffff8000550000 - 0xffffff8008550000 ( 128 MB)
[ 0.000000] vmalloc : 0xffffff8008000000 - 0xffffffbebfff0000 ( 250 GB)
[ 0.000000] .text : 0xffffff8008080000 - 0xffffff8008550000 ( 4928 KB)
[ 0.000000] .rodata : 0xffffff8008550000 - 0xffffff80086a0000 ( 1344 KB)
[ 0.000000] .init : 0xffffff80086a0000 - 0xffffff8008a30000 ( 3648 KB)
[ 0.000000] .data : 0xffffff8008a30000 - 0xffffff8008ab9200 ( 549 KB)
[ 0.000000] .bss : 0xffffff8008ab9200 - 0xffffff8008b0b238 ( 329 KB)
[ 0.000000] fixed : 0xffffffbefe7fd000 - 0xffffffbefec00000 ( 4108 KB)
[ 0.000000] PCI I/O : 0xffffffbefee00000 - 0xffffffbeffe00000 ( 16 MB)
[ 0.000000] vmemmap : 0xffffffbf00000000 - 0xffffffc000000000 ( 4 GB maximum)
[ 0.000000] 0xffffffbf00000000 - 0xffffffbf02000000 ( 32 MB actual)
[ 0.000000] memory : 0xffffffc000000000 - 0xffffffc080000000 ( 2048 MB)
\# cat kernel_page_tables
---[ Modules start ]---
---[ Modules end ]---
---[ vmalloc() Area ]---
0xffffff8008000000-0xffffff8008010000 64K PTE RW NX SHD AF
0xffffff8008015000-0xffffff8008016000 4K PTE RW NX SHD AF
...
2.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
CONFIG_KASAN=n
[ 0.000000] Virtual kernel memory layout:
[ 0.000000] vmalloc : 0xffffff8008000000 - 0xffffffbebfff0000 ( 250 GB)
[ 0.000000] .text : 0xffffff902b280000 - 0xffffff902b750000 ( 4928 KB)
[ 0.000000] .rodata : 0xffffff902b750000 - 0xffffff902b8a0000 ( 1344 KB)
[ 0.000000] .init : 0xffffff902b8a0000 - 0xffffff902bc30000 ( 3648 KB)
[ 0.000000] .data : 0xffffff902bc30000 - 0xffffff902bcb9200 ( 549 KB)
[ 0.000000] .bss : 0xffffff902bcb9200 - 0xffffff902bd0b238 ( 329 KB)
[ 0.000000] fixed : 0xffffffbefe7fd000 - 0xffffffbefec00000 ( 4108 KB)
[ 0.000000] PCI I/O : 0xffffffbefee00000 - 0xffffffbeffe00000 ( 16 MB)
[ 0.000000] vmemmap : 0xffffffbf00000000 - 0xffffffc000000000 ( 4 GB maximum)
[ 0.000000] 0xffffffbf22000000 - 0xffffffbf24000000 ( 32 MB actual)
[ 0.000000] memory : 0xffffffc880000000 - 0xffffffc900000000 ( 2048 MB)
\# cat kernel_page_tables
---[ vmalloc() Area ]---
0xffffff8008000000-0xffffff8008010000 64K PTE RW NX SHD AF
0xffffff8008015000-0xffffff8008016000 4K PTE RW NX SHD AF
0xffffff8008020000-0xffffff8008030000 64K PTE RW NX SHD AF
0xffffff8008031000-0xffffff8008071000 256K PTE RW NX SHD AF
0xffffff800813a000-0xffffff800813d000 12K PTE RW NX SHD AF
...
3.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
CONFIG_KASAN=y
[ 0.000000] Virtual kernel memory layout:
[ 0.000000] kasan : 0xffffff8000000000 - 0xffffff9000000000 ( 64 GB)
[ 0.000000] modules : 0xffffff9000560000 - 0xffffff9008560000 ( 128 MB)
[ 0.000000] vmalloc : 0xffffff9008560000 - 0xffffffbebfff0000 ( 186 GB)
[ 0.000000] .text : 0xffffffa02b280000 - 0xffffffa02b760000 ( 4992 KB)
[ 0.000000] .rodata : 0xffffffa02b760000 - 0xffffffa02b8b0000 ( 1344 KB)
[ 0.000000] .init : 0xffffffa02b8b0000 - 0xffffffa02bc40000 ( 3648 KB)
[ 0.000000] .data : 0xffffffa02bc40000 - 0xffffffa02bcc9200 ( 549 KB)
[ 0.000000] .bss : 0xffffffa02bcc9200 - 0xffffffa02c5342b8 ( 8621 KB)
[ 0.000000] fixed : 0xffffffbefe7fd000 - 0xffffffbefec00000 ( 4108 KB)
[ 0.000000] PCI I/O : 0xffffffbefee00000 - 0xffffffbeffe00000 ( 16 MB)
[ 0.000000] vmemmap : 0xffffffbf00000000 - 0xffffffc000000000 ( 4 GB maximum)
[ 0.000000] 0xffffffbf22000000 - 0xffffffbf24000000 ( 32 MB actual)
[ 0.000000] memory : 0xffffffc880000000 - 0xffffffc900000000 ( 2048 MB)
\# cat kernel_page_tables
---[ Kasan shadow start ]---
0xffffff8000000000-0xffffff82000ac000 8389296K PTE ro NX SHD AF
0xffffff82010ac000-0xffffff8405600000 8459600K PTE ro NX SHD AF
0xffffff8405600000-0xffffff8405a00000 4M PMD RW NX SHD AF
0xffffff8405a00000-0xffffff8800000000 16294M PTE ro NX SHD AF
0xffffff8910000000-0xffffff8920000000 256M PMD RW NX SHD AF
---[ Kasan shadow end ]---
---[ Modules start ]---
---[ Modules end ]---
---[ vmalloc() Area ]---
0xffffff9008560000-0xffffff9008570000 64K PTE RW NX SHD AF
0xffffff9008575000-0xffffff9008576000 4K PTE RW NX SHD AF
0xffffff9008580000-0xffffff9008590000 64K PTE RW NX SHD AF
0xffffff9008591000-0xffffff90085d1000 256K PTE RW NX SHD AF
...
Signed-off-by: Miles Chen <miles.chen@mediatek.com>
---
arch/arm64/include/asm/memory.h | 12 +++++++--
arch/arm64/include/asm/module.h | 6 -----
arch/arm64/include/asm/pgtable.h | 4 +++
arch/arm64/mm/dump.c | 53 ++++++++++++++++++++++++++++++++++++----
arch/arm64/mm/init.c | 6 +++--
5 files changed, 66 insertions(+), 15 deletions(-)
diff --git a/arch/arm64/include/asm/memory.h b/arch/arm64/include/asm/memory.h
index ef39dcb..41b885c 100644
--- a/arch/arm64/include/asm/memory.h
+++ b/arch/arm64/include/asm/memory.h
@@ -68,9 +68,11 @@
(UL(1) << VA_BITS) + 1)
#define PAGE_OFFSET (UL(0xffffffffffffffff) - \
(UL(1) << (VA_BITS - 1)) + 1)
-#define KIMAGE_VADDR (MODULES_END)
+#define KIMAGE_VADDR (STATIC_MODULES_END)
+#define STATIC_MODULES_END (STATIC_MODULES_VADDR + MODULES_VSIZE)
+#define STATIC_MODULES_VADDR (VA_START + KASAN_SHADOW_SIZE)
#define MODULES_END (MODULES_VADDR + MODULES_VSIZE)
-#define MODULES_VADDR (VA_START + KASAN_SHADOW_SIZE)
+#define MODULES_VADDR ((unsigned long)module_alloc_base)
#define MODULES_VSIZE (SZ_128M)
#define VMEMMAP_START (PAGE_OFFSET - VMEMMAP_SIZE)
#define PCI_IO_END (VMEMMAP_START - SZ_2M)
@@ -138,6 +140,12 @@
#include <linux/bitops.h>
#include <linux/mmdebug.h>
+#ifdef CONFIG_RANDOMIZE_BASE
+extern u64 module_alloc_base;
+#else
+#define module_alloc_base ((u64)_etext - MODULES_VSIZE)
+#endif
+
extern s64 memstart_addr;
/* PHYS_OFFSET - the physical address of the start of memory. */
#define PHYS_OFFSET ({ VM_BUG_ON(memstart_addr & 1); memstart_addr; })
diff --git a/arch/arm64/include/asm/module.h b/arch/arm64/include/asm/module.h
index 19bd976..fdf0db4 100644
--- a/arch/arm64/include/asm/module.h
+++ b/arch/arm64/include/asm/module.h
@@ -39,10 +39,4 @@ struct mod_arch_specific {
u64 module_emit_plt_entry(struct module *mod, void *loc, const Elf64_Rela *rela,
Elf64_Sym *sym);
-#ifdef CONFIG_RANDOMIZE_BASE
-extern u64 module_alloc_base;
-#else
-#define module_alloc_base ((u64)_etext - MODULES_VSIZE)
-#endif
-
#endif /* __ASM_MODULE_H */
diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index 6eae342..6e309aa 100644
--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -30,7 +30,11 @@
* VMALLOC_END: extends to the available space below vmmemmap, PCI I/O space
* and fixed mappings
*/
+#ifdef CONFIG_KASAN
#define VMALLOC_START (MODULES_END)
+#else
+#define VMALLOC_START (STATIC_MODULES_END)
+#endif
#define VMALLOC_END (PAGE_OFFSET - PUD_SIZE - VMEMMAP_SIZE - SZ_64K)
#define vmemmap ((struct page *)VMEMMAP_START - (memstart_addr >> PAGE_SHIFT))
diff --git a/arch/arm64/mm/dump.c b/arch/arm64/mm/dump.c
index ca74a2a..18be771 100644
--- a/arch/arm64/mm/dump.c
+++ b/arch/arm64/mm/dump.c
@@ -29,15 +29,36 @@
#include <asm/pgtable-hwdef.h>
#include <asm/ptdump.h>
-static const struct addr_marker address_markers[] = {
+enum marker {
+#ifdef CONFIG_KASAN
+ E_KASAN_SHADOW_START,
+ E_KASAN_SHADOW_END,
+#endif
+ E_MODULES_VADDR,
+ E_MODULES_END,
+ E_VMALLOC_START,
+ E_VMALLOC_END,
+ E_FIXADDR_START,
+ E_FIXADDR_TOP,
+ E_PCI_IO_START,
+ E_PCI_IO_END,
+#ifdef CONFIG_SPARSEMEM_VMEMMAP
+ E_VMEMMAP_START,
+ E_VMEMMAP_END,
+#endif
+ E_PAGE_OFFSET,
+ E_NR_MARKER,
+};
+
+static struct addr_marker address_markers[] = {
#ifdef CONFIG_KASAN
{ KASAN_SHADOW_START, "Kasan shadow start" },
{ KASAN_SHADOW_END, "Kasan shadow end" },
#endif
- { MODULES_VADDR, "Modules start" },
- { MODULES_END, "Modules end" },
- { VMALLOC_START, "vmalloc() Area" },
- { VMALLOC_END, "vmalloc() End" },
+ { -1, "Modules start" },
+ { -1, "Modules end" },
+ { -1, "vmalloc() Area" },
+ { -1, "vmalloc() End" },
{ FIXADDR_START, "Fixmap start" },
{ FIXADDR_TOP, "Fixmap end" },
{ PCI_IO_START, "PCI I/O start" },
@@ -362,10 +383,32 @@ void ptdump_walk_pgd(struct seq_file *m, struct ptdump_info *info)
note_page(&st, 0, 0, 0);
}
+static void fixup_markers(void)
+{
+ int i;
+
+ address_markers[E_MODULES_VADDR].start_address = MODULES_VADDR;
+ address_markers[E_MODULES_END].start_address = MODULES_END;
+ address_markers[E_VMALLOC_START].start_address = VMALLOC_START;
+ address_markers[E_VMALLOC_END].start_address = VMALLOC_END;
+
+ if (MODULES_VADDR < VMALLOC_START) {
+ address_markers[E_MODULES_END].start_address =
+ (MODULES_END < VMALLOC_START) ?
+ MODULES_END : VMALLOC_START;
+ } else {
+ /* modules is contains in vamlloc area, don't show them */
+ for (i = E_MODULES_VADDR; i <= E_NR_MARKER - 2; i++)
+ address_markers[i] = address_markers[i + 2];
+ }
+}
+
static void ptdump_initialize(void)
{
unsigned i, j;
+ fixup_markers();
+
for (i = 0; i < ARRAY_SIZE(pg_level); i++)
if (pg_level[i].bits)
for (j = 0; j < pg_level[i].num; j++)
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index 5960bef..be77b26 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -609,8 +609,10 @@ void __init mem_init(void)
pr_notice(" kasan : 0x%16lx - 0x%16lx (%6ld GB)\n",
MLG(KASAN_SHADOW_START, KASAN_SHADOW_END));
#endif
- pr_notice(" modules : 0x%16lx - 0x%16lx (%6ld MB)\n",
- MLM(MODULES_VADDR, MODULES_END));
+ /* print modules info when it's not fully mapped within vamlloc */
+ if (MODULES_VADDR < VMALLOC_START)
+ pr_notice(" modules : 0x%16lx - 0x%16lx (%6ld MB)\n",
+ MLM(MODULES_VADDR, MODULES_END));
pr_notice(" vmalloc : 0x%16lx - 0x%16lx (%6ld GB)\n",
MLG(VMALLOC_START, VMALLOC_END));
pr_notice(" .text : 0x%p" " - 0x%p" " (%6ld KB)\n",
--
1.9.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* [RFC/PATCH v3] arm64: define MODULES_VADDR by module_alloc_base
2017-08-30 11:46 [RFC/PATCH v3] arm64: define MODULES_VADDR by module_alloc_base Miles Chen
@ 2017-10-17 14:43 ` Will Deacon
2017-10-19 11:08 ` Miles Chen
0 siblings, 1 reply; 3+ messages in thread
From: Will Deacon @ 2017-10-17 14:43 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Aug 30, 2017 at 07:46:33PM +0800, Miles Chen wrote:
> After the kernel ASLR, the module virtual address is moved to
> [module_alloc_base, module_alloc_base + MODULES_VSIZE).
> However, the MODULES_VADDR is still defined as a constant and functions
> like is_vmalloc_or_module_addr() or dump functions will not able to
> use correct module range information.
>
> The patch omits modules information in virtual kenrel memory layout if
> the modules area is fully mapped whitin vmalloc area.
>
> Use module_alloc_base to define MODULES_VADDR. I tested the patch under
> three different conditions:
> 1.CONFIG_RANDOMIZE_BASE=y, seed=0
> CONFIG_KASAN=n
> 2.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
> CONFIG_KASAN=n
> 3.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
> CONFIG_KASAN=y
>
> test log:
[...]
> diff --git a/arch/arm64/mm/dump.c b/arch/arm64/mm/dump.c
> index ca74a2a..18be771 100644
> --- a/arch/arm64/mm/dump.c
> +++ b/arch/arm64/mm/dump.c
> @@ -29,15 +29,36 @@
> #include <asm/pgtable-hwdef.h>
> #include <asm/ptdump.h>
>
> -static const struct addr_marker address_markers[] = {
> +enum marker {
> +#ifdef CONFIG_KASAN
> + E_KASAN_SHADOW_START,
> + E_KASAN_SHADOW_END,
> +#endif
> + E_MODULES_VADDR,
> + E_MODULES_END,
> + E_VMALLOC_START,
> + E_VMALLOC_END,
> + E_FIXADDR_START,
> + E_FIXADDR_TOP,
> + E_PCI_IO_START,
> + E_PCI_IO_END,
> +#ifdef CONFIG_SPARSEMEM_VMEMMAP
> + E_VMEMMAP_START,
> + E_VMEMMAP_END,
> +#endif
> + E_PAGE_OFFSET,
> + E_NR_MARKER,
> +};
> +
> +static struct addr_marker address_markers[] = {
> #ifdef CONFIG_KASAN
> { KASAN_SHADOW_START, "Kasan shadow start" },
> { KASAN_SHADOW_END, "Kasan shadow end" },
> #endif
> - { MODULES_VADDR, "Modules start" },
> - { MODULES_END, "Modules end" },
> - { VMALLOC_START, "vmalloc() Area" },
> - { VMALLOC_END, "vmalloc() End" },
> + { -1, "Modules start" },
> + { -1, "Modules end" },
> + { -1, "vmalloc() Area" },
> + { -1, "vmalloc() End" },
> { FIXADDR_START, "Fixmap start" },
> { FIXADDR_TOP, "Fixmap end" },
> { PCI_IO_START, "PCI I/O start" },
> @@ -362,10 +383,32 @@ void ptdump_walk_pgd(struct seq_file *m, struct ptdump_info *info)
> note_page(&st, 0, 0, 0);
> }
>
> +static void fixup_markers(void)
> +{
> + int i;
> +
> + address_markers[E_MODULES_VADDR].start_address = MODULES_VADDR;
> + address_markers[E_MODULES_END].start_address = MODULES_END;
> + address_markers[E_VMALLOC_START].start_address = VMALLOC_START;
> + address_markers[E_VMALLOC_END].start_address = VMALLOC_END;
> +
> + if (MODULES_VADDR < VMALLOC_START) {
> + address_markers[E_MODULES_END].start_address =
> + (MODULES_END < VMALLOC_START) ?
> + MODULES_END : VMALLOC_START;
> + } else {
> + /* modules is contains in vamlloc area, don't show them */
> + for (i = E_MODULES_VADDR; i <= E_NR_MARKER - 2; i++)
> + address_markers[i] = address_markers[i + 2];
> + }
> +}
This all seems a bit over-engineered to me and we end up having to maintain
enum marker in conjunction with the address_markers array. Fixing the array
up at runtime also worries me slightly, because we end up with the last two
entries being duplicated. That doesn't seem to hurt for now, but it's weird
and I can imagine it causing problems in the future.
Is there not a simpler fix here?
Will
^ permalink raw reply [flat|nested] 3+ messages in thread
* [RFC/PATCH v3] arm64: define MODULES_VADDR by module_alloc_base
2017-10-17 14:43 ` Will Deacon
@ 2017-10-19 11:08 ` Miles Chen
0 siblings, 0 replies; 3+ messages in thread
From: Miles Chen @ 2017-10-19 11:08 UTC (permalink / raw)
To: linux-arm-kernel
On Tue, 2017-10-17 at 15:43 +0100, Will Deacon wrote:
> On Wed, Aug 30, 2017 at 07:46:33PM +0800, Miles Chen wrote:
> > After the kernel ASLR, the module virtual address is moved to
> > [module_alloc_base, module_alloc_base + MODULES_VSIZE).
> > However, the MODULES_VADDR is still defined as a constant and functions
> > like is_vmalloc_or_module_addr() or dump functions will not able to
> > use correct module range information.
> >
> > The patch omits modules information in virtual kenrel memory layout if
> > the modules area is fully mapped whitin vmalloc area.
> >
> > Use module_alloc_base to define MODULES_VADDR. I tested the patch under
> > three different conditions:
> > 1.CONFIG_RANDOMIZE_BASE=y, seed=0
> > CONFIG_KASAN=n
> > 2.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
> > CONFIG_KASAN=n
> > 3.CONFIG_RANDOMIZE_BASE=y, seed=0x2304909023333333
> > CONFIG_KASAN=y
> >
> > test log:
>
> [...]
>
> > diff --git a/arch/arm64/mm/dump.c b/arch/arm64/mm/dump.c
> > index ca74a2a..18be771 100644
> > --- a/arch/arm64/mm/dump.c
> > +++ b/arch/arm64/mm/dump.c
> > @@ -29,15 +29,36 @@
> > #include <asm/pgtable-hwdef.h>
> > #include <asm/ptdump.h>
> >
> > -static const struct addr_marker address_markers[] = {
> > +enum marker {
> > +#ifdef CONFIG_KASAN
> > + E_KASAN_SHADOW_START,
> > + E_KASAN_SHADOW_END,
> > +#endif
> > + E_MODULES_VADDR,
> > + E_MODULES_END,
> > + E_VMALLOC_START,
> > + E_VMALLOC_END,
> > + E_FIXADDR_START,
> > + E_FIXADDR_TOP,
> > + E_PCI_IO_START,
> > + E_PCI_IO_END,
> > +#ifdef CONFIG_SPARSEMEM_VMEMMAP
> > + E_VMEMMAP_START,
> > + E_VMEMMAP_END,
> > +#endif
> > + E_PAGE_OFFSET,
> > + E_NR_MARKER,
> > +};
> > +
> > +static struct addr_marker address_markers[] = {
> > #ifdef CONFIG_KASAN
> > { KASAN_SHADOW_START, "Kasan shadow start" },
> > { KASAN_SHADOW_END, "Kasan shadow end" },
> > #endif
> > - { MODULES_VADDR, "Modules start" },
> > - { MODULES_END, "Modules end" },
> > - { VMALLOC_START, "vmalloc() Area" },
> > - { VMALLOC_END, "vmalloc() End" },
> > + { -1, "Modules start" },
> > + { -1, "Modules end" },
> > + { -1, "vmalloc() Area" },
> > + { -1, "vmalloc() End" },
> > { FIXADDR_START, "Fixmap start" },
> > { FIXADDR_TOP, "Fixmap end" },
> > { PCI_IO_START, "PCI I/O start" },
> > @@ -362,10 +383,32 @@ void ptdump_walk_pgd(struct seq_file *m, struct ptdump_info *info)
> > note_page(&st, 0, 0, 0);
> > }
> >
> > +static void fixup_markers(void)
> > +{
> > + int i;
> > +
> > + address_markers[E_MODULES_VADDR].start_address = MODULES_VADDR;
> > + address_markers[E_MODULES_END].start_address = MODULES_END;
> > + address_markers[E_VMALLOC_START].start_address = VMALLOC_START;
> > + address_markers[E_VMALLOC_END].start_address = VMALLOC_END;
> > +
> > + if (MODULES_VADDR < VMALLOC_START) {
> > + address_markers[E_MODULES_END].start_address =
> > + (MODULES_END < VMALLOC_START) ?
> > + MODULES_END : VMALLOC_START;
> > + } else {
> > + /* modules is contains in vamlloc area, don't show them */
> > + for (i = E_MODULES_VADDR; i <= E_NR_MARKER - 2; i++)
> > + address_markers[i] = address_markers[i + 2];
> > + }
> > +}
>
> This all seems a bit over-engineered to me and we end up having to maintain
> enum marker in conjunction with the address_markers array. Fixing the array
> up at runtime also worries me slightly, because we end up with the last two
> entries being duplicated. That doesn't seem to hurt for now, but it's weird
> and I can imagine it causing problems in the future.
>
> Is there not a simpler fix here?
>
> Will
Sorry for the late reply.
How about using static defined STATIC_MODULES_VADDR, and
STATIC_MODULES_END, so we do not need to fixed the address_markers[].
- { MODULES_VADDR, "Modules start" },
- { MODULES_END, "Modules end" },
+ { STATIC_MODULES_VADDR, "Static modules start" },
+ { STATIC_MODULES_END, "Static modules end" },
and for the modules:
1. set the [STATIC_MODULES_VADDR,STATIC_MODULES_END) as the default
module area (e.g., KASAN=n + KASLR=y with no random seed; KASLR=y and
KASN=y or KASLR=n)
2. with KASAN=n + KASLR=y with valid random seed, kernel modules are
loaded to vmalloc area. We can still the STATIC_MODULES_* entries in the
address_markers[]. We should find nothing between
[STATIC_MODULES_VADDR,STATIC_MODULES_END) in this case.
Miles
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2017-10-19 11:08 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-08-30 11:46 [RFC/PATCH v3] arm64: define MODULES_VADDR by module_alloc_base Miles Chen
2017-10-17 14:43 ` Will Deacon
2017-10-19 11:08 ` Miles Chen
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).