OpenE2K/qemu-e2k
13
4
Fork 0
Code Issues 4 Pull Requests Projects 2 Releases Activity

cpu-exec: simplify jump cache management

Browse Source 
Unless I'm missing something egregious, the jmp cache is only every
populated with a valid entry by the same thread that reads the cache.
Therefore, the contents of any valid entry are always consistent and
there is no need for any acquire/release magic.

Indeed ->tb has to be accessed with atomics, because concurrent
invalidations would otherwise cause data races.  But ->pc is only ever
accessed by one thread, and accesses to ->tb and ->pc within tb_lookup
can never race with another tb_lookup.  While the TranslationBlock
(especially the flags) could be modified by a concurrent invalidation,
store-release and load-acquire operations on the cache entry would
not add any additional ordering beyond what you get from performing
the accesses within a single thread.

Because of this, there is really nothing to win in splitting the CF_PCREL
and !CF_PCREL paths.  It is easier to just always use the ->pc field in
the jump cache.

I noticed this while working on splitting commit 8ed558ec0c
("accel/tcg: Introduce TARGET_TB_PCREL", 2022-10-04) into multiple
pieces, for the sake of finding a more fine-grained bisection
result for https://gitlab.com/qemu-project/qemu/-/issues/2092.
It does not (and does not intend to) fix that issue; therefore
it may make sense to not commit it until the root cause
of issue #2092 is found.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240122153409.351959-1-pbonzini@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This commit is contained in:
Paolo Bonzini 2024-01-22 16:34:09 +01:00 committed by Richard Henderson
parent 7a1dc45af5
commit d157e540ed
2 changed files with 28 additions and 46 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

66
accel/tcg/cpu-exec.c
View File

@ -253,43 +253,29 @@ static inline TranslationBlock *tb_lookup(CPUState *cpu, vaddr pc,
hash = tb_jmp_cache_hash_func(pc);
jc = cpu->tb_jmp_cache;
if (cflags & CF_PCREL) {
/* Use acquire to ensure current load of pc from jc. */
tb = qatomic_load_acquire(&jc->array[hash].tb);
if (likely(tb &&
jc->array[hash].pc == pc &&
tb->cs_base == cs_base &&
tb->flags == flags &&
tb_cflags(tb) == cflags)) {
return tb;
}
tb = tb_htable_lookup(cpu, pc, cs_base, flags, cflags);
if (tb == NULL) {
return NULL;
}
jc->array[hash].pc = pc;
/* Ensure pc is written first. */
qatomic_store_release(&jc->array[hash].tb, tb);
} else {
/* Use rcu_read to ensure current load of pc from *tb. */
tb = qatomic_rcu_read(&jc->array[hash].tb);
if (likely(tb &&
tb->pc == pc &&
tb->cs_base == cs_base &&
tb->flags == flags &&
tb_cflags(tb) == cflags)) {
return tb;
}
tb = tb_htable_lookup(cpu, pc, cs_base, flags, cflags);
if (tb == NULL) {
return NULL;
}
/* Use the pc value already stored in tb->pc. */
qatomic_set(&jc->array[hash].tb, tb);
tb = qatomic_read(&jc->array[hash].tb);
if (likely(tb &&
jc->array[hash].pc == pc &&
tb->cs_base == cs_base &&
tb->flags == flags &&
tb_cflags(tb) == cflags)) {
goto hit;
}
tb = tb_htable_lookup(cpu, pc, cs_base, flags, cflags);
if (tb == NULL) {
return NULL;
}
jc->array[hash].pc = pc;
qatomic_set(&jc->array[hash].tb, tb);
hit:
/*
* As long as tb is not NULL, the contents are consistent. Therefore,
* the virtual PC has to match for non-CF_PCREL translations.
*/
assert((tb_cflags(tb) & CF_PCREL) || tb->pc == pc);
return tb;
}
@ -1012,14 +998,8 @@ cpu_exec_loop(CPUState *cpu, SyncClocks *sc)
*/
h = tb_jmp_cache_hash_func(pc);
jc = cpu->tb_jmp_cache;
if (cflags & CF_PCREL) {
jc->array[h].pc = pc;
/* Ensure pc is written first. */
qatomic_store_release(&jc->array[h].tb, tb);
} else {
/* Use the pc value already stored in tb->pc. */
qatomic_set(&jc->array[h].tb, tb);
}
jc->array[h].pc = pc;
qatomic_set(&jc->array[h].tb, tb);
}
#ifndef CONFIG_USER_ONLY

8
accel/tcg/tb-jmp-cache.h
View File

@ -13,9 +13,11 @@
#define TB_JMP_CACHE_SIZE (1 << TB_JMP_CACHE_BITS)
/*
* Accessed in parallel; all accesses to 'tb' must be atomic.
* For CF_PCREL, accesses to 'pc' must be protected by a
* load_acquire/store_release to 'tb'.
* Invalidated in parallel; all accesses to 'tb' must be atomic.
* A valid entry is read/written by a single CPU, therefore there is
* no need for qatomic_rcu_read() and pc is always consistent with a
* non-NULL value of 'tb'. Strictly speaking pc is only needed for
* CF_PCREL, but it's used always for simplicity.
*/
struct CPUJumpCache {
struct rcu_head rcu;