1
Just one bugfix patch for this rc:
1
This bug seemed worth fixing for 8.0 since we need an rc4 anyway:
2
we were using uninitialized data for the guarded bit when
3
combining stage 1 and stage 2 attrs.
2
4
3
The following changes since commit ca5f3d4df1b47d7f66a109cdb504e83dfd7ec433:
5
thanks
6
-- PMM
4
7
5
Merge tag 'pull-la-20220808' of https://gitlab.com/rth7680/qemu into staging (2022-08-08 19:51:12 -0700)
8
The following changes since commit 08dede07030973c1053868bc64de7e10bfa02ad6:
9
10
Merge tag 'pull-ppc-20230409' of https://github.com/legoater/qemu into staging (2023-04-10 11:47:52 +0100)
6
11
7
are available in the Git repository at:
12
are available in the Git repository at:
8
13
9
https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20220809
14
https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20230410
10
15
11
for you to fetch changes up to c7f26ded6d5065e4116f630f6a490b55f6c5f58e:
16
for you to fetch changes up to 8539dc00552e8ea60420856fc1262c8299bc6308:
12
17
13
icount: Take iothread lock when running QEMU timers (2022-08-09 10:55:14 +0100)
18
target/arm: Copy guarded bit in combine_cacheattrs (2023-04-10 14:31:40 +0100)
14
19
15
----------------------------------------------------------------
20
----------------------------------------------------------------
16
target-arm queue:
21
target-arm: Fix bug where we weren't initializing
17
* icount: Take iothread lock when running QEMU timers
22
guarded bit state when combining S1/S2 attrs
18
23
19
----------------------------------------------------------------
24
----------------------------------------------------------------
20
Peter Maydell (1):
25
Richard Henderson (2):
21
icount: Take iothread lock when running QEMU timers
26
target/arm: PTE bit GP only applies to stage1
27
target/arm: Copy guarded bit in combine_cacheattrs
22
28
23
accel/tcg/tcg-accel-ops-icount.c | 6 ++++++
29
target/arm/ptw.c | 11 ++++++-----
24
1 file changed, 6 insertions(+)
30
1 file changed, 6 insertions(+), 5 deletions(-)
diff view generated by jsdifflib
New patch
1
From: Richard Henderson <richard.henderson@linaro.org>
1
2
3
Only perform the extract of GP during the stage1 walk.
4
5
Reported-by: Peter Maydell <peter.maydell@linaro.org>
6
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
7
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
8
Message-id: 20230407185149.3253946-2-richard.henderson@linaro.org
9
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10
---
11
target/arm/ptw.c | 10 +++++-----
12
1 file changed, 5 insertions(+), 5 deletions(-)
13
14
diff --git a/target/arm/ptw.c b/target/arm/ptw.c
15
index XXXXXXX..XXXXXXX 100644
16
--- a/target/arm/ptw.c
17
+++ b/target/arm/ptw.c
18
@@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, S1Translate *ptw,
19
result->f.attrs.secure = false;
20
}
21
22
- /* When in aarch64 mode, and BTI is enabled, remember GP in the TLB. */
23
- if (aarch64 && cpu_isar_feature(aa64_bti, cpu)) {
24
- result->f.guarded = extract64(attrs, 50, 1); /* GP */
25
- }
26
-
27
if (regime_is_stage2(mmu_idx)) {
28
result->cacheattrs.is_s2_format = true;
29
result->cacheattrs.attrs = extract32(attrs, 2, 4);
30
@@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, S1Translate *ptw,
31
assert(attrindx <= 7);
32
result->cacheattrs.is_s2_format = false;
33
result->cacheattrs.attrs = extract64(mair, attrindx * 8, 8);
34
+
35
+ /* When in aarch64 mode, and BTI is enabled, remember GP in the TLB. */
36
+ if (aarch64 && cpu_isar_feature(aa64_bti, cpu)) {
37
+ result->f.guarded = extract64(attrs, 50, 1); /* GP */
38
+ }
39
}
40
41
/*
42
--
43
2.34.1
diff view generated by jsdifflib
1
The function icount_prepare_for_run() is called with the iothread
1
From: Richard Henderson <richard.henderson@linaro.org>
2
unlocked, but it can call icount_notify_aio_contexts() which will
3
run qemu timer handlers. Those are supposed to be run only with
4
the iothread lock held, so take the lock while we do that.
5
2
6
Since icount mode runs everything on a single thread anyway,
3
The guarded bit comes from the stage1 walk.
7
not holding the lock is likely mostly not going to introduce
8
races, but it can cause us to trip over assertions that we
9
do hold the lock, such as the one reported in issue 1130.
10
4
11
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1130
5
Fixes: Coverity CID 1507929
6
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
7
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
8
Message-id: 20230407185149.3253946-3-richard.henderson@linaro.org
12
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
9
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
13
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
14
Tested-by: Pavel Dovgalyuk <Pavel.Dovgalyuk@ispras.ru>
15
Message-id: 20220801164527.3134765-1-peter.maydell@linaro.org
16
---
10
---
17
accel/tcg/tcg-accel-ops-icount.c | 6 ++++++
11
target/arm/ptw.c | 1 +
18
1 file changed, 6 insertions(+)
12
1 file changed, 1 insertion(+)
19
13
20
diff --git a/accel/tcg/tcg-accel-ops-icount.c b/accel/tcg/tcg-accel-ops-icount.c
14
diff --git a/target/arm/ptw.c b/target/arm/ptw.c
21
index XXXXXXX..XXXXXXX 100644
15
index XXXXXXX..XXXXXXX 100644
22
--- a/accel/tcg/tcg-accel-ops-icount.c
16
--- a/target/arm/ptw.c
23
+++ b/accel/tcg/tcg-accel-ops-icount.c
17
+++ b/target/arm/ptw.c
24
@@ -XXX,XX +XXX,XX @@ void icount_prepare_for_run(CPUState *cpu)
18
@@ -XXX,XX +XXX,XX @@ static ARMCacheAttrs combine_cacheattrs(uint64_t hcr,
25
replay_mutex_lock();
19
26
20
assert(!s1.is_s2_format);
27
if (cpu->icount_budget == 0) {
21
ret.is_s2_format = false;
28
+ /*
22
+ ret.guarded = s1.guarded;
29
+ * We're called without the iothread lock, so must take it while
23
30
+ * we're calling timer handlers.
24
if (s1.attrs == 0xf0) {
31
+ */
25
tagged = true;
32
+ qemu_mutex_lock_iothread();
33
icount_notify_aio_contexts();
34
+ qemu_mutex_unlock_iothread();
35
}
36
}
37
38
--
26
--
39
2.25.1
27
2.34.1
diff view generated by jsdifflib