[Qemu-devel] [PATCH v5 0/6] per-TLB lock

[Qemu-devel] [PATCH v5 0/6] per-TLB lock

[Emilio G. Cota]

v4: https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg01421.html

Changes since v4:

- Add two patches to remove early calls to tlb_flush.

You can fetch the series from:
  https://github.com/cota/qemu/tree/tlb-lock-v5

Thanks,

		Emilio

[Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Emilio G. Cota]

As far as I can tell tlb_flush does not need to be called
this early. tlb_flush is eventually called after the CPU
has been realized.

This change paves the way to the introduction of tlb_init,
which will be called from cpu_exec_realizefn.

Signed-off-by: Emilio G. Cota <cota@braap.org>
---
 target/alpha/cpu.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index b08078e7fc..a953897fcc 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -201,7 +201,6 @@ static void alpha_cpu_initfn(Object *obj)
     CPUAlphaState *env = &cpu->env;
 
     cs->env_ptr = env;
-    tlb_flush(cs);
 
     env->lock_addr = -1;
 #if defined(CONFIG_USER_ONLY)
-- 
2.17.1

Re: [Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Peter Maydell]

On 9 October 2018 at 18:45, Emilio G. Cota <cota@braap.org> wrote:
> As far as I can tell tlb_flush does not need to be called
> this early. tlb_flush is eventually called after the CPU
> has been realized.
>
> This change paves the way to the introduction of tlb_init,
> which will be called from cpu_exec_realizefn.
>
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
>  target/alpha/cpu.c | 1 -
>  1 file changed, 1 deletion(-)
>
> diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
> index b08078e7fc..a953897fcc 100644
> --- a/target/alpha/cpu.c
> +++ b/target/alpha/cpu.c
> @@ -201,7 +201,6 @@ static void alpha_cpu_initfn(Object *obj)
>      CPUAlphaState *env = &cpu->env;
>
>      cs->env_ptr = env;
> -    tlb_flush(cs);
>
>      env->lock_addr = -1;
>  #if defined(CONFIG_USER_ONLY)
> --
> 2.17.1

Definitely agreed that we don't want to tlb_flush in the
target cpu initfn.


What's the codepath by which tlb_flush gets called on
cpu reset? I had a quick look but couldn't find it...

(The other dubious-looking bit of flushing in the
target/alpha code is the code that generates calls
to tb_flush()... we have very few calls to tb_flush
outside the 'core' code and I suspect they could all
be avoided.)

thanks
-- PMM

Re: [Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Emilio G. Cota]

On Tue, Oct 09, 2018 at 18:55:30 +0100, Peter Maydell wrote:
> On 9 October 2018 at 18:45, Emilio G. Cota <cota@braap.org> wrote:
(snip)
> > @@ -201,7 +201,6 @@ static void alpha_cpu_initfn(Object *obj)
> >      CPUAlphaState *env = &cpu->env;
> >
> >      cs->env_ptr = env;
> > -    tlb_flush(cs);
> >
> >      env->lock_addr = -1;
> >  #if defined(CONFIG_USER_ONLY)
> > --
> > 2.17.1
> 
> Definitely agreed that we don't want to tlb_flush in the
> target cpu initfn.
> 
> 
> What's the codepath by which tlb_flush gets called on
> cpu reset? I had a quick look but couldn't find it...

>From cpu.c:

static void cpu_common_reset(CPUState *cpu)
{
    CPUClass *cc = CPU_GET_CLASS(cpu);
...
    if (tcg_enabled()) {
        cpu_tb_jmp_cache_clear(cpu);

        tcg_flush_softmmu_tlb(cpu);
    }
}

tcg_flush_softmmu_tlb is defined in translate-all.c:

/* This is a wrapper for common code that can not use CONFIG_SOFTMMU */
void tcg_flush_softmmu_tlb(CPUState *cs)
{
#ifdef CONFIG_SOFTMMU
    tlb_flush(cs);
#endif
}

> (The other dubious-looking bit of flushing in the
> target/alpha code is the code that generates calls
> to tb_flush()... we have very few calls to tb_flush
> outside the 'core' code and I suspect they could all
> be avoided.)

If the comment below is accurate, seems fair enough.
tb_flush is only called from target/alpha through a helper,
generated by:

        /* PALBR */
        tcg_gen_st_i64(vb, cpu_env, offsetof(CPUAlphaState, palbr));
        /* Changing the PAL base register implies un-chaining all of the TBs
           that ended with a CALL_PAL.  Since the base register usually only
           changes during boot, flushing everything works well.  */
        gen_helper_tb_flush(cpu_env);
        return DISAS_PC_STALE;

Thanks,

		Emilio

Re: [Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Peter Maydell]

On 9 October 2018 at 19:20, Emilio G. Cota <cota@braap.org> wrote:
> On Tue, Oct 09, 2018 at 18:55:30 +0100, Peter Maydell wrote:
>>
>> What's the codepath by which tlb_flush gets called on
>> cpu reset? I had a quick look but couldn't find it...
>
> From cpu.c:
>
> static void cpu_common_reset(CPUState *cpu)
> {
>     CPUClass *cc = CPU_GET_CLASS(cpu);
> ...
>     if (tcg_enabled()) {
>         cpu_tb_jmp_cache_clear(cpu);
>
>         tcg_flush_softmmu_tlb(cpu);
>     }
> }
>
> tcg_flush_softmmu_tlb is defined in translate-all.c:
>
> /* This is a wrapper for common code that can not use CONFIG_SOFTMMU */
> void tcg_flush_softmmu_tlb(CPUState *cs)
> {
> #ifdef CONFIG_SOFTMMU
>     tlb_flush(cs);
> #endif
> }

Ah, thank you. I missed this because of the indirection
via tcg_flush_softmmu_tlb().

>> (The other dubious-looking bit of flushing in the
>> target/alpha code is the code that generates calls
>> to tb_flush()... we have very few calls to tb_flush
>> outside the 'core' code and I suspect they could all
>> be avoided.)
>
> If the comment below is accurate, seems fair enough.
> tb_flush is only called from target/alpha through a helper,
> generated by:
>
>         /* PALBR */
>         tcg_gen_st_i64(vb, cpu_env, offsetof(CPUAlphaState, palbr));
>         /* Changing the PAL base register implies un-chaining all of the TBs
>            that ended with a CALL_PAL.  Since the base register usually only
>            changes during boot, flushing everything works well.  */
>         gen_helper_tb_flush(cpu_env);
>         return DISAS_PC_STALE;

Mmm, it works (though would it work if the CPU wasn't the
only one in the system?). I just have a reflexive dislike
of design approaches used by only one thing -- I tend to
like to smooth them out so there's more consistency...

thanks
-- PMM

Re: [Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Richard Henderson]

On 10/9/18 11:25 AM, Peter Maydell wrote:
>>         /* PALBR */
>>         tcg_gen_st_i64(vb, cpu_env, offsetof(CPUAlphaState, palbr));
>>         /* Changing the PAL base register implies un-chaining all of the TBs
>>            that ended with a CALL_PAL.  Since the base register usually only
>>            changes during boot, flushing everything works well.  */
>>         gen_helper_tb_flush(cpu_env);
>>         return DISAS_PC_STALE;
> 
> Mmm, it works (though would it work if the CPU wasn't the
> only one in the system?). I just have a reflexive dislike
> of design approaches used by only one thing -- I tend to
> like to smooth them out so there's more consistency...

It could be less important now that we have goto_ptr.
If you really think I should get rid of it, I will.


r~

Re: [Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Alex Bennée]

Emilio G. Cota <cota@braap.org> writes:

> As far as I can tell tlb_flush does not need to be called
> this early. tlb_flush is eventually called after the CPU
> has been realized.
>
> This change paves the way to the introduction of tlb_init,
> which will be called from cpu_exec_realizefn.
>
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
>  target/alpha/cpu.c | 1 -
>  1 file changed, 1 deletion(-)
>
> diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
> index b08078e7fc..a953897fcc 100644
> --- a/target/alpha/cpu.c
> +++ b/target/alpha/cpu.c
> @@ -201,7 +201,6 @@ static void alpha_cpu_initfn(Object *obj)
>      CPUAlphaState *env = &cpu->env;
>
>      cs->env_ptr = env;
> -    tlb_flush(cs);
>
>      env->lock_addr = -1;
>  #if defined(CONFIG_USER_ONLY)

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>

--
Alex Bennée

Re: [Qemu-devel] [PATCH v5 1/6] target/alpha: remove tlb_flush from alpha_cpu_initfn

[Richard Henderson]

On 10/9/18 10:45 AM, Emilio G. Cota wrote:
> As far as I can tell tlb_flush does not need to be called
> this early. tlb_flush is eventually called after the CPU
> has been realized.
> 
> This change paves the way to the introduction of tlb_init,
> which will be called from cpu_exec_realizefn.
> 
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
>  target/alpha/cpu.c | 1 -
>  1 file changed, 1 deletion(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[Qemu-devel] [PATCH v5 2/6] target/unicore32: remove tlb_flush from uc32_init_fn

[Emilio G. Cota]

As far as I can tell tlb_flush does not need to be called
this early. tlb_flush is eventually called after the CPU
has been realized.

This change paves the way to the introduction of tlb_init,
which will be called from cpu_exec_realizefn.

Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Signed-off-by: Emilio G. Cota <cota@braap.org>
---
 target/unicore32/cpu.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
index 68f978d80b..2b49d1ca40 100644
--- a/target/unicore32/cpu.c
+++ b/target/unicore32/cpu.c
@@ -116,8 +116,6 @@ static void uc32_cpu_initfn(Object *obj)
     env->uncached_asr = ASR_MODE_PRIV;
     env->regs[31] = 0x03000000;
 #endif
-
-    tlb_flush(cs);
 }
 
 static const VMStateDescription vmstate_uc32_cpu = {
-- 
2.17.1

Re: [Qemu-devel] [PATCH v5 2/6] target/unicore32: remove tlb_flush from uc32_init_fn

[Alex Bennée]

Emilio G. Cota <cota@braap.org> writes:

> As far as I can tell tlb_flush does not need to be called
> this early. tlb_flush is eventually called after the CPU
> has been realized.
>
> This change paves the way to the introduction of tlb_init,
> which will be called from cpu_exec_realizefn.
>
> Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
>  target/unicore32/cpu.c | 2 --
>  1 file changed, 2 deletions(-)
>
> diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
> index 68f978d80b..2b49d1ca40 100644
> --- a/target/unicore32/cpu.c
> +++ b/target/unicore32/cpu.c
> @@ -116,8 +116,6 @@ static void uc32_cpu_initfn(Object *obj)
>      env->uncached_asr = ASR_MODE_PRIV;
>      env->regs[31] = 0x03000000;
>  #endif
> -
> -    tlb_flush(cs);
>  }
>
>  static const VMStateDescription vmstate_uc32_cpu = {

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>

--
Alex Bennée

Re: [Qemu-devel] [PATCH v5 2/6] target/unicore32: remove tlb_flush from uc32_init_fn

[Richard Henderson]

On 10/9/18 10:45 AM, Emilio G. Cota wrote:
> As far as I can tell tlb_flush does not need to be called
> this early. tlb_flush is eventually called after the CPU
> has been realized.
> 
> This change paves the way to the introduction of tlb_init,
> which will be called from cpu_exec_realizefn.
> 
> Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
>  target/unicore32/cpu.c | 2 --
>  1 file changed, 2 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[Qemu-devel] [PATCH v5 3/6] exec: introduce tlb_init

[Emilio G. Cota]

Paves the way for the addition of a per-TLB lock.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
---
 include/exec/exec-all.h | 8 ++++++++
 accel/tcg/cputlb.c      | 4 ++++
 exec.c                  | 1 +
 3 files changed, 13 insertions(+)

diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h
index 5f78125582..815e5b1e83 100644
--- a/include/exec/exec-all.h
+++ b/include/exec/exec-all.h
@@ -99,6 +99,11 @@ void cpu_address_space_init(CPUState *cpu, int asidx,
 
 #if !defined(CONFIG_USER_ONLY) && defined(CONFIG_TCG)
 /* cputlb.c */
+/**
+ * tlb_init - initialize a CPU's TLB
+ * @cpu: CPU whose TLB should be initialized
+ */
+void tlb_init(CPUState *cpu);
 /**
  * tlb_flush_page:
  * @cpu: CPU whose TLB should be flushed
@@ -258,6 +263,9 @@ void tlb_set_page(CPUState *cpu, target_ulong vaddr,
 void probe_write(CPUArchState *env, target_ulong addr, int size, int mmu_idx,
                  uintptr_t retaddr);
 #else
+static inline void tlb_init(CPUState *cpu)
+{
+}
 static inline void tlb_flush_page(CPUState *cpu, target_ulong addr)
 {
 }
diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
index f4702ce91f..502eea2850 100644
--- a/accel/tcg/cputlb.c
+++ b/accel/tcg/cputlb.c
@@ -73,6 +73,10 @@ QEMU_BUILD_BUG_ON(sizeof(target_ulong) > sizeof(run_on_cpu_data));
 QEMU_BUILD_BUG_ON(NB_MMU_MODES > 16);
 #define ALL_MMUIDX_BITS ((1 << NB_MMU_MODES) - 1)
 
+void tlb_init(CPUState *cpu)
+{
+}
+
 /* flush_all_helper: run fn across all cpus
  *
  * If the wait flag is set then the src cpu's helper will be queued as
diff --git a/exec.c b/exec.c
index d0821e69aa..4fd831ef06 100644
--- a/exec.c
+++ b/exec.c
@@ -965,6 +965,7 @@ void cpu_exec_realizefn(CPUState *cpu, Error **errp)
         tcg_target_initialized = true;
         cc->tcg_initialize();
     }
+    tlb_init(cpu);
 
 #ifndef CONFIG_USER_ONLY
     if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
-- 
2.17.1

Re: [Qemu-devel] [PATCH v5 3/6] exec: introduce tlb_init

[Richard Henderson]

On 10/9/18 10:45 AM, Emilio G. Cota wrote:
> Paves the way for the addition of a per-TLB lock.
> 
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
> Signed-off-by: Emilio G. Cota <cota@braap.org>
> ---
>  include/exec/exec-all.h | 8 ++++++++
>  accel/tcg/cputlb.c      | 4 ++++
>  exec.c                  | 1 +
>  3 files changed, 13 insertions(+)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[Qemu-devel] [PATCH v5 4/6] cputlb: fix assert_cpu_is_self macro

[Emilio G. Cota]

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
---
 accel/tcg/cputlb.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
index 502eea2850..f6b388c961 100644
--- a/accel/tcg/cputlb.c
+++ b/accel/tcg/cputlb.c
@@ -58,9 +58,9 @@
     } \
 } while (0)
 
-#define assert_cpu_is_self(this_cpu) do {                         \
+#define assert_cpu_is_self(cpu) do {                              \
         if (DEBUG_TLB_GATE) {                                     \
-            g_assert(!cpu->created || qemu_cpu_is_self(cpu));     \
+            g_assert(!(cpu)->created || qemu_cpu_is_self(cpu));   \
         }                                                         \
     } while (0)
 
-- 
2.17.1

[Qemu-devel] [PATCH v5 5/6] cputlb: serialize tlb updates with env->tlb_lock

[Emilio G. Cota]

Currently we rely on atomic operations for cross-CPU invalidations.
There are two cases that these atomics miss: cross-CPU invalidations
can race with either (1) vCPU threads flushing their TLB, which
happens via memset, or (2) vCPUs calling tlb_reset_dirty on their TLB,
which updates .addr_write with a regular store. This results in
undefined behaviour, since we're mixing regular and atomic ops
on concurrent accesses.

Fix it by using tlb_lock, a per-vCPU lock. All updaters of tlb_table
and the corresponding victim cache now hold the lock.
The readers that do not hold tlb_lock must use atomic reads when
reading .addr_write, since this field can be updated by other threads;
the conversion to atomic reads is done in the next patch.

Note that an alternative fix would be to expand the use of atomic ops.
However, in the case of TLB flushes this would have a huge performance
impact, since (1) TLB flushes can happen very frequently and (2) we
currently use a full memory barrier to flush each TLB entry, and a TLB
has many entries. Instead, acquiring the lock is barely slower than a
full memory barrier since it is uncontended, and with a single lock
acquisition we can flush the entire TLB.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
---
 include/exec/cpu-defs.h |   3 +
 accel/tcg/cputlb.c      | 155 ++++++++++++++++++++++------------------
 2 files changed, 87 insertions(+), 71 deletions(-)

diff --git a/include/exec/cpu-defs.h b/include/exec/cpu-defs.h
index a171ffc1a4..4ff62f32bf 100644
--- a/include/exec/cpu-defs.h
+++ b/include/exec/cpu-defs.h
@@ -24,6 +24,7 @@
 #endif
 
 #include "qemu/host-utils.h"
+#include "qemu/thread.h"
 #include "qemu/queue.h"
 #ifdef CONFIG_TCG
 #include "tcg-target.h"
@@ -142,6 +143,8 @@ typedef struct CPUIOTLBEntry {
 
 #define CPU_COMMON_TLB \
     /* The meaning of the MMU modes is defined in the target code. */   \
+    /* tlb_lock serializes updates to tlb_table and tlb_v_table */      \
+    QemuSpin tlb_lock;                                                  \
     CPUTLBEntry tlb_table[NB_MMU_MODES][CPU_TLB_SIZE];                  \
     CPUTLBEntry tlb_v_table[NB_MMU_MODES][CPU_VTLB_SIZE];               \
     CPUIOTLBEntry iotlb[NB_MMU_MODES][CPU_TLB_SIZE];                    \
diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
index f6b388c961..c2a6190674 100644
--- a/accel/tcg/cputlb.c
+++ b/accel/tcg/cputlb.c
@@ -75,6 +75,9 @@ QEMU_BUILD_BUG_ON(NB_MMU_MODES > 16);
 
 void tlb_init(CPUState *cpu)
 {
+    CPUArchState *env = cpu->env_ptr;
+
+    qemu_spin_init(&env->tlb_lock);
 }
 
 /* flush_all_helper: run fn across all cpus
@@ -129,8 +132,17 @@ static void tlb_flush_nocheck(CPUState *cpu)
     atomic_set(&env->tlb_flush_count, env->tlb_flush_count + 1);
     tlb_debug("(count: %zu)\n", tlb_flush_count());
 
+    /*
+     * tlb_table/tlb_v_table updates from any thread must hold tlb_lock.
+     * However, updates from the owner thread (as is the case here; see the
+     * above assert_cpu_is_self) do not need atomic_set because all reads
+     * that do not hold the lock are performed by the same owner thread.
+     */
+    qemu_spin_lock(&env->tlb_lock);
     memset(env->tlb_table, -1, sizeof(env->tlb_table));
     memset(env->tlb_v_table, -1, sizeof(env->tlb_v_table));
+    qemu_spin_unlock(&env->tlb_lock);
+
     cpu_tb_jmp_cache_clear(cpu);
 
     env->vtlb_index = 0;
@@ -182,6 +194,7 @@ static void tlb_flush_by_mmuidx_async_work(CPUState *cpu, run_on_cpu_data data)
 
     tlb_debug("start: mmu_idx:0x%04lx\n", mmu_idx_bitmask);
 
+    qemu_spin_lock(&env->tlb_lock);
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
 
         if (test_bit(mmu_idx, &mmu_idx_bitmask)) {
@@ -191,6 +204,7 @@ static void tlb_flush_by_mmuidx_async_work(CPUState *cpu, run_on_cpu_data data)
             memset(env->tlb_v_table[mmu_idx], -1, sizeof(env->tlb_v_table[0]));
         }
     }
+    qemu_spin_unlock(&env->tlb_lock);
 
     cpu_tb_jmp_cache_clear(cpu);
 
@@ -247,19 +261,24 @@ static inline bool tlb_hit_page_anyprot(CPUTLBEntry *tlb_entry,
            tlb_hit_page(tlb_entry->addr_code, page);
 }
 
-static inline void tlb_flush_entry(CPUTLBEntry *tlb_entry, target_ulong page)
+/* Called with tlb_lock held */
+static inline void tlb_flush_entry_locked(CPUTLBEntry *tlb_entry,
+                                          target_ulong page)
 {
     if (tlb_hit_page_anyprot(tlb_entry, page)) {
         memset(tlb_entry, -1, sizeof(*tlb_entry));
     }
 }
 
-static inline void tlb_flush_vtlb_page(CPUArchState *env, int mmu_idx,
-                                       target_ulong page)
+/* Called with tlb_lock held */
+static inline void tlb_flush_vtlb_page_locked(CPUArchState *env, int mmu_idx,
+                                              target_ulong page)
 {
     int k;
+
+    assert_cpu_is_self(ENV_GET_CPU(env));
     for (k = 0; k < CPU_VTLB_SIZE; k++) {
-        tlb_flush_entry(&env->tlb_v_table[mmu_idx][k], page);
+        tlb_flush_entry_locked(&env->tlb_v_table[mmu_idx][k], page);
     }
 }
 
@@ -286,10 +305,12 @@ static void tlb_flush_page_async_work(CPUState *cpu, run_on_cpu_data data)
 
     addr &= TARGET_PAGE_MASK;
     i = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
+    qemu_spin_lock(&env->tlb_lock);
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
-        tlb_flush_entry(&env->tlb_table[mmu_idx][i], addr);
-        tlb_flush_vtlb_page(env, mmu_idx, addr);
+        tlb_flush_entry_locked(&env->tlb_table[mmu_idx][i], addr);
+        tlb_flush_vtlb_page_locked(env, mmu_idx, addr);
     }
+    qemu_spin_unlock(&env->tlb_lock);
 
     tb_flush_jmp_cache(cpu, addr);
 }
@@ -326,12 +347,14 @@ static void tlb_flush_page_by_mmuidx_async_work(CPUState *cpu,
     tlb_debug("page:%d addr:"TARGET_FMT_lx" mmu_idx:0x%lx\n",
               page, addr, mmu_idx_bitmap);
 
+    qemu_spin_lock(&env->tlb_lock);
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
         if (test_bit(mmu_idx, &mmu_idx_bitmap)) {
-            tlb_flush_entry(&env->tlb_table[mmu_idx][page], addr);
-            tlb_flush_vtlb_page(env, mmu_idx, addr);
+            tlb_flush_entry_locked(&env->tlb_table[mmu_idx][page], addr);
+            tlb_flush_vtlb_page_locked(env, mmu_idx, addr);
         }
     }
+    qemu_spin_unlock(&env->tlb_lock);
 
     tb_flush_jmp_cache(cpu, addr);
 }
@@ -454,72 +477,44 @@ void tlb_unprotect_code(ram_addr_t ram_addr)
  * most usual is detecting writes to code regions which may invalidate
  * generated code.
  *
- * Because we want other vCPUs to respond to changes straight away we
- * update the te->addr_write field atomically. If the TLB entry has
- * been changed by the vCPU in the mean time we skip the update.
+ * Other vCPUs might be reading their TLBs during guest execution, so we update
+ * te->addr_write with atomic_set. We don't need to worry about this for
+ * oversized guests as MTTCG is disabled for them.
  *
- * As this function uses atomic accesses we also need to ensure
- * updates to tlb_entries follow the same access rules. We don't need
- * to worry about this for oversized guests as MTTCG is disabled for
- * them.
+ * Called with tlb_lock held.
  */
-
-static void tlb_reset_dirty_range(CPUTLBEntry *tlb_entry, uintptr_t start,
-                           uintptr_t length)
+static void tlb_reset_dirty_range_locked(CPUTLBEntry *tlb_entry,
+                                         uintptr_t start, uintptr_t length)
 {
-#if TCG_OVERSIZED_GUEST
     uintptr_t addr = tlb_entry->addr_write;
 
     if ((addr & (TLB_INVALID_MASK | TLB_MMIO | TLB_NOTDIRTY)) == 0) {
         addr &= TARGET_PAGE_MASK;
         addr += tlb_entry->addend;
         if ((addr - start) < length) {
+#if TCG_OVERSIZED_GUEST
             tlb_entry->addr_write |= TLB_NOTDIRTY;
-        }
-    }
 #else
-    /* paired with atomic_mb_set in tlb_set_page_with_attrs */
-    uintptr_t orig_addr = atomic_mb_read(&tlb_entry->addr_write);
-    uintptr_t addr = orig_addr;
-
-    if ((addr & (TLB_INVALID_MASK | TLB_MMIO | TLB_NOTDIRTY)) == 0) {
-        addr &= TARGET_PAGE_MASK;
-        addr += atomic_read(&tlb_entry->addend);
-        if ((addr - start) < length) {
-            uintptr_t notdirty_addr = orig_addr | TLB_NOTDIRTY;
-            atomic_cmpxchg(&tlb_entry->addr_write, orig_addr, notdirty_addr);
+            atomic_set(&tlb_entry->addr_write,
+                       tlb_entry->addr_write | TLB_NOTDIRTY);
+#endif
         }
     }
-#endif
 }
 
-/* For atomic correctness when running MTTCG we need to use the right
- * primitives when copying entries */
-static inline void copy_tlb_helper(CPUTLBEntry *d, CPUTLBEntry *s,
-                                   bool atomic_set)
+/*
+ * Called with tlb_lock held.
+ * Called only from the vCPU context, i.e. the TLB's owner thread.
+ */
+static inline void copy_tlb_helper_locked(CPUTLBEntry *d, const CPUTLBEntry *s)
 {
-#if TCG_OVERSIZED_GUEST
     *d = *s;
-#else
-    if (atomic_set) {
-        d->addr_read = s->addr_read;
-        d->addr_code = s->addr_code;
-        atomic_set(&d->addend, atomic_read(&s->addend));
-        /* Pairs with flag setting in tlb_reset_dirty_range */
-        atomic_mb_set(&d->addr_write, atomic_read(&s->addr_write));
-    } else {
-        d->addr_read = s->addr_read;
-        d->addr_write = atomic_read(&s->addr_write);
-        d->addr_code = s->addr_code;
-        d->addend = atomic_read(&s->addend);
-    }
-#endif
 }
 
 /* This is a cross vCPU call (i.e. another vCPU resetting the flags of
- * the target vCPU). As such care needs to be taken that we don't
- * dangerously race with another vCPU update. The only thing actually
- * updated is the target TLB entry ->addr_write flags.
+ * the target vCPU).
+ * We must take tlb_lock to avoid racing with another vCPU update. The only
+ * thing actually updated is the target TLB entry ->addr_write flags.
  */
 void tlb_reset_dirty(CPUState *cpu, ram_addr_t start1, ram_addr_t length)
 {
@@ -528,22 +523,26 @@ void tlb_reset_dirty(CPUState *cpu, ram_addr_t start1, ram_addr_t length)
     int mmu_idx;
 
     env = cpu->env_ptr;
+    qemu_spin_lock(&env->tlb_lock);
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
         unsigned int i;
 
         for (i = 0; i < CPU_TLB_SIZE; i++) {
-            tlb_reset_dirty_range(&env->tlb_table[mmu_idx][i],
-                                  start1, length);
+            tlb_reset_dirty_range_locked(&env->tlb_table[mmu_idx][i], start1,
+                                         length);
         }
 
         for (i = 0; i < CPU_VTLB_SIZE; i++) {
-            tlb_reset_dirty_range(&env->tlb_v_table[mmu_idx][i],
-                                  start1, length);
+            tlb_reset_dirty_range_locked(&env->tlb_v_table[mmu_idx][i], start1,
+                                         length);
         }
     }
+    qemu_spin_unlock(&env->tlb_lock);
 }
 
-static inline void tlb_set_dirty1(CPUTLBEntry *tlb_entry, target_ulong vaddr)
+/* Called with tlb_lock held */
+static inline void tlb_set_dirty1_locked(CPUTLBEntry *tlb_entry,
+                                         target_ulong vaddr)
 {
     if (tlb_entry->addr_write == (vaddr | TLB_NOTDIRTY)) {
         tlb_entry->addr_write = vaddr;
@@ -562,16 +561,18 @@ void tlb_set_dirty(CPUState *cpu, target_ulong vaddr)
 
     vaddr &= TARGET_PAGE_MASK;
     i = (vaddr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
+    qemu_spin_lock(&env->tlb_lock);
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
-        tlb_set_dirty1(&env->tlb_table[mmu_idx][i], vaddr);
+        tlb_set_dirty1_locked(&env->tlb_table[mmu_idx][i], vaddr);
     }
 
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
         int k;
         for (k = 0; k < CPU_VTLB_SIZE; k++) {
-            tlb_set_dirty1(&env->tlb_v_table[mmu_idx][k], vaddr);
+            tlb_set_dirty1_locked(&env->tlb_v_table[mmu_idx][k], vaddr);
         }
     }
+    qemu_spin_unlock(&env->tlb_lock);
 }
 
 /* Our TLB does not support large pages, so remember the area covered by
@@ -658,9 +659,6 @@ void tlb_set_page_with_attrs(CPUState *cpu, target_ulong vaddr,
         addend = (uintptr_t)memory_region_get_ram_ptr(section->mr) + xlat;
     }
 
-    /* Make sure there's no cached translation for the new page.  */
-    tlb_flush_vtlb_page(env, mmu_idx, vaddr_page);
-
     code_address = address;
     iotlb = memory_region_section_get_iotlb(cpu, section, vaddr_page,
                                             paddr_page, xlat, prot, &address);
@@ -668,6 +666,18 @@ void tlb_set_page_with_attrs(CPUState *cpu, target_ulong vaddr,
     index = (vaddr_page >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
     te = &env->tlb_table[mmu_idx][index];
 
+    /*
+     * Hold the TLB lock for the rest of the function. We could acquire/release
+     * the lock several times in the function, but it is faster to amortize the
+     * acquisition cost by acquiring it just once. Note that this leads to
+     * a longer critical section, but this is not a concern since the TLB lock
+     * is unlikely to be contended.
+     */
+    qemu_spin_lock(&env->tlb_lock);
+
+    /* Make sure there's no cached translation for the new page.  */
+    tlb_flush_vtlb_page_locked(env, mmu_idx, vaddr_page);
+
     /*
      * Only evict the old entry to the victim tlb if it's for a
      * different page; otherwise just overwrite the stale data.
@@ -677,7 +687,7 @@ void tlb_set_page_with_attrs(CPUState *cpu, target_ulong vaddr,
         CPUTLBEntry *tv = &env->tlb_v_table[mmu_idx][vidx];
 
         /* Evict the old entry into the victim tlb.  */
-        copy_tlb_helper(tv, te, true);
+        copy_tlb_helper_locked(tv, te);
         env->iotlb_v[mmu_idx][vidx] = env->iotlb[mmu_idx][index];
     }
 
@@ -729,9 +739,8 @@ void tlb_set_page_with_attrs(CPUState *cpu, target_ulong vaddr,
         }
     }
 
-    /* Pairs with flag setting in tlb_reset_dirty_range */
-    copy_tlb_helper(te, &tn, true);
-    /* atomic_mb_set(&te->addr_write, write_address); */
+    copy_tlb_helper_locked(te, &tn);
+    qemu_spin_unlock(&env->tlb_lock);
 }
 
 /* Add a new TLB entry, but without specifying the memory
@@ -895,6 +904,8 @@ static bool victim_tlb_hit(CPUArchState *env, size_t mmu_idx, size_t index,
                            size_t elt_ofs, target_ulong page)
 {
     size_t vidx;
+
+    assert_cpu_is_self(ENV_GET_CPU(env));
     for (vidx = 0; vidx < CPU_VTLB_SIZE; ++vidx) {
         CPUTLBEntry *vtlb = &env->tlb_v_table[mmu_idx][vidx];
         target_ulong cmp = *(target_ulong *)((uintptr_t)vtlb + elt_ofs);
@@ -903,9 +914,11 @@ static bool victim_tlb_hit(CPUArchState *env, size_t mmu_idx, size_t index,
             /* Found entry in victim tlb, swap tlb and iotlb.  */
             CPUTLBEntry tmptlb, *tlb = &env->tlb_table[mmu_idx][index];
 
-            copy_tlb_helper(&tmptlb, tlb, false);
-            copy_tlb_helper(tlb, vtlb, true);
-            copy_tlb_helper(vtlb, &tmptlb, true);
+            qemu_spin_lock(&env->tlb_lock);
+            copy_tlb_helper_locked(&tmptlb, tlb);
+            copy_tlb_helper_locked(tlb, vtlb);
+            copy_tlb_helper_locked(vtlb, &tmptlb);
+            qemu_spin_unlock(&env->tlb_lock);
 
             CPUIOTLBEntry tmpio, *io = &env->iotlb[mmu_idx][index];
             CPUIOTLBEntry *vio = &env->iotlb_v[mmu_idx][vidx];
-- 
2.17.1

[Qemu-devel] [PATCH v5 6/6] cputlb: read CPUTLBEntry.addr_write atomically

[Emilio G. Cota]

Updates can come from other threads, so readers that do not
take tlb_lock must use atomic_read to avoid undefined
behaviour (UB).

This and the previous commit result on average in no performance loss,
as the following experiments (run on an Intel i7-6700K CPU @ 4.00GHz)
show.

1. aarch64 bootup+shutdown test:

- Before:
 Performance counter stats for 'taskset -c 0 ../img/aarch64/die.sh' (10 runs):

       7487.087786      task-clock (msec)         #    0.998 CPUs utilized            ( +-  0.12% )
    31,574,905,303      cycles                    #    4.217 GHz                      ( +-  0.12% )
    57,097,908,812      instructions              #    1.81  insns per cycle          ( +-  0.08% )
    10,255,415,367      branches                  # 1369.747 M/sec                    ( +-  0.08% )
       173,278,962      branch-misses             #    1.69% of all branches          ( +-  0.18% )

       7.504481349 seconds time elapsed                                          ( +-  0.14% )

- After:
 Performance counter stats for 'taskset -c 0 ../img/aarch64/die.sh' (10 runs):

       7462.441328      task-clock (msec)         #    0.998 CPUs utilized            ( +-  0.07% )
    31,478,476,520      cycles                    #    4.218 GHz                      ( +-  0.07% )
    57,017,330,084      instructions              #    1.81  insns per cycle          ( +-  0.05% )
    10,251,929,667      branches                  # 1373.804 M/sec                    ( +-  0.05% )
       173,023,787      branch-misses             #    1.69% of all branches          ( +-  0.11% )

       7.474970463 seconds time elapsed                                          ( +-  0.07% )

2. SPEC06int:
                                              SPEC06int (test set)
                                           [Y axis: Speedup over master]
  1.15 +-+----+------+------+------+------+------+-------+------+------+------+------+------+------+----+-+
       |                                                                                                  |
   1.1 +-+.................................+++.............................+  tlb-lock-v2 (m+++x)       +-+
       |                                +++ |                   +++        tlb-lock-v3 (spinl|ck)         |
       |                    +++          |  |     +++    +++     |                           |            |
  1.05 +-+....+++...........####.........|####.+++.|......|.....###....+++...........+++....###.........+-+
       |      ###         ++#| #         |# |# ***### +++### +++#+#     |     +++     |     #|#    ###    |
     1 +-+++***+#++++####+++#++#++++++++++#++#+*+*++#++++#+#+****+#++++###++++###++++###++++#+#++++#+#+++-+
       |    *+* #    #++# ***  #   #### ***  # * *++# ****+# *| * # ****|#   |# #    #|#    #+#    # #    |
  0.95 +-+..*.*.#....#..#.*|*..#...#..#.*|*..#.*.*..#.*|.*.#.*++*.#.*++*+#.****.#....#+#....#.#..++#.#..+-+
       |    * * #    #  # *|*  #   #  # *|*  # * *  # *++* # *  * # *  * # * |* #  ++# #    # #  *** #    |
       |    * * #  ++#  # *+*  #   #  # *|*  # * *  # *  * # *  * # *  * # *++* # **** #  ++# #  * * #    |
   0.9 +-+..*.*.#...|#..#.*.*..#.++#..#.*|*..#.*.*..#.*..*.#.*..*.#.*..*.#.*..*.#.*.|*.#...|#.#..*.*.#..+-+
       |    * * #  ***  # * *  #  |#  # *+*  # * *  # *  * # *  * # *  * # *  * # *++* #   |# #  * * #    |
  0.85 +-+..*.*.#..*|*..#.*.*..#.***..#.*.*..#.*.*..#.*..*.#.*..*.#.*..*.#.*..*.#.*..*.#.****.#..*.*.#..+-+
       |    * * #  *+*  # * *  # *|*  # * *  # * *  # *  * # *  * # *  * # *  * # *  * # * |* #  * * #    |
       |    * * #  * *  # * *  # *+*  # * *  # * *  # *  * # *  * # *  * # *  * # *  * # * |* #  * * #    |
   0.8 +-+..*.*.#..*.*..#.*.*..#.*.*..#.*.*..#.*.*..#.*..*.#.*..*.#.*..*.#.*..*.#.*..*.#.*++*.#..*.*.#..+-+
       |    * * #  * *  # * *  # * *  # * *  # * *  # *  * # *  * # *  * # *  * # *  * # *  * #  * * #    |
  0.75 +-+--***##--***###-***###-***###-***###-***###-****##-****##-****##-****##-****##-****##--***##--+-+
 400.perlben401.bzip2403.gcc429.m445.gob456.hmme45462.libqua464.h26471.omnet473483.xalancbmkgeomean

  png: https://imgur.com/a/BHzpPTW

Notes:
- tlb-lock-v2 corresponds to an implementation with a mutex.
- tlb-lock-v3 is the current patch series, i.e. with a spinlock
  and a single lock acquisition in tlb_set_page_with_attrs.

Signed-off-by: Emilio G. Cota <cota@braap.org>
---
 accel/tcg/softmmu_template.h     | 16 ++++++++++------
 include/exec/cpu_ldst.h          |  2 +-
 include/exec/cpu_ldst_template.h |  2 +-
 accel/tcg/cputlb.c               | 15 +++++++++------
 4 files changed, 21 insertions(+), 14 deletions(-)

diff --git a/accel/tcg/softmmu_template.h b/accel/tcg/softmmu_template.h
index f060a693d4..1e50263871 100644
--- a/accel/tcg/softmmu_template.h
+++ b/accel/tcg/softmmu_template.h
@@ -277,7 +277,8 @@ void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
 {
     unsigned mmu_idx = get_mmuidx(oi);
     int index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-    target_ulong tlb_addr = env->tlb_table[mmu_idx][index].addr_write;
+    target_ulong tlb_addr =
+        atomic_read(&env->tlb_table[mmu_idx][index].addr_write);
     unsigned a_bits = get_alignment_bits(get_memop(oi));
     uintptr_t haddr;
 
@@ -292,7 +293,8 @@ void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
             tlb_fill(ENV_GET_CPU(env), addr, DATA_SIZE, MMU_DATA_STORE,
                      mmu_idx, retaddr);
         }
-        tlb_addr = env->tlb_table[mmu_idx][index].addr_write & ~TLB_INVALID_MASK;
+        tlb_addr = atomic_read(&env->tlb_table[mmu_idx][index].addr_write) &
+            ~TLB_INVALID_MASK;
     }
 
     /* Handle an IO access.  */
@@ -321,7 +323,7 @@ void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
            cannot evict the first.  */
         page2 = (addr + DATA_SIZE) & TARGET_PAGE_MASK;
         index2 = (page2 >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-        tlb_addr2 = env->tlb_table[mmu_idx][index2].addr_write;
+        tlb_addr2 = atomic_read(&env->tlb_table[mmu_idx][index2].addr_write);
         if (!tlb_hit_page(tlb_addr2, page2)
             && !VICTIM_TLB_HIT(addr_write, page2)) {
             tlb_fill(ENV_GET_CPU(env), page2, DATA_SIZE, MMU_DATA_STORE,
@@ -354,7 +356,8 @@ void helper_be_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
 {
     unsigned mmu_idx = get_mmuidx(oi);
     int index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-    target_ulong tlb_addr = env->tlb_table[mmu_idx][index].addr_write;
+    target_ulong tlb_addr =
+        atomic_read(&env->tlb_table[mmu_idx][index].addr_write);
     unsigned a_bits = get_alignment_bits(get_memop(oi));
     uintptr_t haddr;
 
@@ -369,7 +372,8 @@ void helper_be_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
             tlb_fill(ENV_GET_CPU(env), addr, DATA_SIZE, MMU_DATA_STORE,
                      mmu_idx, retaddr);
         }
-        tlb_addr = env->tlb_table[mmu_idx][index].addr_write & ~TLB_INVALID_MASK;
+        tlb_addr = atomic_read(&env->tlb_table[mmu_idx][index].addr_write) &
+            ~TLB_INVALID_MASK;
     }
 
     /* Handle an IO access.  */
@@ -398,7 +402,7 @@ void helper_be_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
            cannot evict the first.  */
         page2 = (addr + DATA_SIZE) & TARGET_PAGE_MASK;
         index2 = (page2 >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-        tlb_addr2 = env->tlb_table[mmu_idx][index2].addr_write;
+        tlb_addr2 = atomic_read(&env->tlb_table[mmu_idx][index2].addr_write);
         if (!tlb_hit_page(tlb_addr2, page2)
             && !VICTIM_TLB_HIT(addr_write, page2)) {
             tlb_fill(ENV_GET_CPU(env), page2, DATA_SIZE, MMU_DATA_STORE,
diff --git a/include/exec/cpu_ldst.h b/include/exec/cpu_ldst.h
index 41ed0526e2..9581587ce1 100644
--- a/include/exec/cpu_ldst.h
+++ b/include/exec/cpu_ldst.h
@@ -426,7 +426,7 @@ static inline void *tlb_vaddr_to_host(CPUArchState *env, abi_ptr addr,
         tlb_addr = tlbentry->addr_read;
         break;
     case 1:
-        tlb_addr = tlbentry->addr_write;
+        tlb_addr = atomic_read(&tlbentry->addr_write);
         break;
     case 2:
         tlb_addr = tlbentry->addr_code;
diff --git a/include/exec/cpu_ldst_template.h b/include/exec/cpu_ldst_template.h
index 4db2302962..ba7a11123c 100644
--- a/include/exec/cpu_ldst_template.h
+++ b/include/exec/cpu_ldst_template.h
@@ -176,7 +176,7 @@ glue(glue(glue(cpu_st, SUFFIX), MEMSUFFIX), _ra)(CPUArchState *env,
     addr = ptr;
     page_index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
     mmu_idx = CPU_MMU_INDEX;
-    if (unlikely(env->tlb_table[mmu_idx][page_index].addr_write !=
+    if (unlikely(atomic_read(&env->tlb_table[mmu_idx][page_index].addr_write) !=
                  (addr & (TARGET_PAGE_MASK | (DATA_SIZE - 1))))) {
         oi = make_memop_idx(SHIFT, mmu_idx);
         glue(glue(helper_ret_st, SUFFIX), MMUSUFFIX)(env, addr, v, oi,
diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
index c2a6190674..200444142a 100644
--- a/accel/tcg/cputlb.c
+++ b/accel/tcg/cputlb.c
@@ -257,7 +257,7 @@ static inline bool tlb_hit_page_anyprot(CPUTLBEntry *tlb_entry,
                                         target_ulong page)
 {
     return tlb_hit_page(tlb_entry->addr_read, page) ||
-           tlb_hit_page(tlb_entry->addr_write, page) ||
+           tlb_hit_page(atomic_read(&tlb_entry->addr_write), page) ||
            tlb_hit_page(tlb_entry->addr_code, page);
 }
 
@@ -859,7 +859,7 @@ static void io_writex(CPUArchState *env, CPUIOTLBEntry *iotlbentry,
         tlb_fill(cpu, addr, size, MMU_DATA_STORE, mmu_idx, retaddr);
 
         index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-        tlb_addr = env->tlb_table[mmu_idx][index].addr_write;
+        tlb_addr = atomic_read(&env->tlb_table[mmu_idx][index].addr_write);
         if (!(tlb_addr & ~(TARGET_PAGE_MASK | TLB_RECHECK))) {
             /* RAM access */
             uintptr_t haddr = addr + env->tlb_table[mmu_idx][index].addend;
@@ -908,7 +908,9 @@ static bool victim_tlb_hit(CPUArchState *env, size_t mmu_idx, size_t index,
     assert_cpu_is_self(ENV_GET_CPU(env));
     for (vidx = 0; vidx < CPU_VTLB_SIZE; ++vidx) {
         CPUTLBEntry *vtlb = &env->tlb_v_table[mmu_idx][vidx];
-        target_ulong cmp = *(target_ulong *)((uintptr_t)vtlb + elt_ofs);
+        /* elt_ofs might correspond to .addr_write, so use atomic_read */
+        target_ulong cmp =
+            atomic_read((target_ulong *)((uintptr_t)vtlb + elt_ofs));
 
         if (cmp == page) {
             /* Found entry in victim tlb, swap tlb and iotlb.  */
@@ -980,7 +982,8 @@ void probe_write(CPUArchState *env, target_ulong addr, int size, int mmu_idx,
                  uintptr_t retaddr)
 {
     int index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-    target_ulong tlb_addr = env->tlb_table[mmu_idx][index].addr_write;
+    target_ulong tlb_addr =
+        atomic_read(&env->tlb_table[mmu_idx][index].addr_write);
 
     if (!tlb_hit(tlb_addr, addr)) {
         /* TLB entry is for a different page */
@@ -1000,7 +1003,7 @@ static void *atomic_mmu_lookup(CPUArchState *env, target_ulong addr,
     size_t mmu_idx = get_mmuidx(oi);
     size_t index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
     CPUTLBEntry *tlbe = &env->tlb_table[mmu_idx][index];
-    target_ulong tlb_addr = tlbe->addr_write;
+    target_ulong tlb_addr = atomic_read(&tlbe->addr_write);
     TCGMemOp mop = get_memop(oi);
     int a_bits = get_alignment_bits(mop);
     int s_bits = mop & MO_SIZE;
@@ -1031,7 +1034,7 @@ static void *atomic_mmu_lookup(CPUArchState *env, target_ulong addr,
             tlb_fill(ENV_GET_CPU(env), addr, 1 << s_bits, MMU_DATA_STORE,
                      mmu_idx, retaddr);
         }
-        tlb_addr = tlbe->addr_write & ~TLB_INVALID_MASK;
+        tlb_addr = atomic_read(&tlbe->addr_write) & ~TLB_INVALID_MASK;
     }
 
     /* Notice an IO access or a needs-MMU-lookup access */
-- 
2.17.1

Re: [Qemu-devel] [PATCH v5 0/6] per-TLB lock

[Richard Henderson]

On 10/9/18 10:45 AM, Emilio G. Cota wrote:
> v4: https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg01421.html
> 
> Changes since v4:
> 
> - Add two patches to remove early calls to tlb_flush.
> 
> You can fetch the series from:
>   https://github.com/cota/qemu/tree/tlb-lock-v5

Queued, thanks.


r~