* [PATCH memory-model 0/14] Updates to the formal memory model
@ 2018-07-16 18:05 Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 01/14] tools/memory-model: Add litmus test for full multicopy atomicity Paul E. McKenney
` (13 more replies)
0 siblings, 14 replies; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, palmer, sj38.park,
mark.rutland, j.neuschaefer, yauheni.kaliuta
Hello!
This series contains updates to the Linux kernel's formal memory model
in tools/memory-model, along with corresponding changes in documentation
and Linux-kernel code. These patches are ready for inclusion into -tip.
1. Add a litmus test for full multi-copy atomicity.
2. Fix the name of the ISA2+pooncelock+pooncelock+pombonce litmus test.
3. Add Daniel Lustig as an LKMM reviewer, with focus on RISC-V,
courtesy of Palmer Dabbelt.
4. Update Korean translation to fix broken DMA vs. MMIO ordering
example, courtesy of SeongJae Park with edits suggested by
Byungchul Park.
5-6. Remove the now-obsolete ACCESS_ONCE() from recipes and the model
itself, courtesy of Mark Rutland.
7. Make regression-test scripts executable.
8. Describe atomic_set as a write operation, courtesy of Jonathan
Neuschäfer.
9. Add informal LKMM documentation to MAINTAINERS.
10. Use smp_mb() in wake_woken_function(), courtesy of Andrea Parri.
11. Clarify requirements for smp_mb__after_spinlock(), courtesy of
Andrea Parri.
12. Update wake_up() and friends' memory-barrier guarantees,
courtesy of Andrea Parri.
13. Fix "smb_wmb()" typo (should be "smp_wmb()"), courtesy of
Yauheni Kaliuta.
14. Rename litmus tests to comply to norm7, courtesy of Andrea Parri.
Thanx, Paul
------------------------------------------------------------------------
Documentation/core-api/atomic_ops.rst | 2
Documentation/memory-barriers.txt | 43 +++---
Documentation/translations/ko_KR/memory-barriers.txt | 22 +--
MAINTAINERS | 6
include/linux/sched.h | 4
include/linux/spinlock.h | 53 +++++--
kernel/sched/completion.c | 8 -
kernel/sched/core.c | 71 ++++------
kernel/sched/wait.c | 55 +++----
tools/memory-model/Documentation/explanation.txt | 2
tools/memory-model/Documentation/recipes.txt | 12 -
tools/memory-model/README | 20 +-
tools/memory-model/linux-kernel.bell | 2
tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus | 2
tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus | 2
tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus | 2
tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus | 2
tools/memory-model/litmus-tests/R+fencembonceonces.litmus | 2
tools/memory-model/litmus-tests/README | 25 ++-
tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus | 2
tools/memory-model/litmus-tests/SB+fencembonceonces.litmus | 2
tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus | 32 ++++
tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus | 2
tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus | 2
tools/memory-model/scripts/checkalllitmus.sh | 2
tools/memory-model/scripts/checklitmus.sh | 2
26 files changed, 223 insertions(+), 156 deletions(-)
^ permalink raw reply [flat|nested] 28+ messages in thread
* [PATCH memory-model 01/14] tools/memory-model: Add litmus test for full multicopy atomicity
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:25 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 02/14] tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name Paul E. McKenney
` (12 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Paul E. McKenney
This commit adds a litmus test suggested by Alan Stern that is forbidden
on fully multicopy atomic systems, but allowed on other-multicopy and
on non-multicopy atomic systems. For reference, s390 is fully multicopy
atomic, x86 and ARMv8 are other-multicopy atomic, and ARMv7 and powerpc
are non-multicopy atomic.
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
---
tools/memory-model/litmus-tests/README | 9 ++++++
.../SB+rfionceonce-poonceonces.litmus | 32 +++++++++++++++++++
2 files changed, 41 insertions(+)
create mode 100644 tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus
diff --git a/tools/memory-model/litmus-tests/README b/tools/memory-model/litmus-tests/README
index 17eb9a8c222d..00140aaf58b7 100644
--- a/tools/memory-model/litmus-tests/README
+++ b/tools/memory-model/litmus-tests/README
@@ -111,6 +111,15 @@ SB+mbonceonces.litmus
SB+poonceonces.litmus
As above, but without the smp_mb() invocations.
+SB+rfionceonce-poonceonces.litmus
+ This litmus test demonstrates that LKMM is not fully multicopy
+ atomic. (Neither is it other multicopy atomic.) This litmus test
+ also demonstrates the "locations" debugging aid, which designates
+ additional registers and locations to be printed out in the dump
+ of final states in the herd7 output. Without the "locations"
+ statement, only those registers and locations mentioned in the
+ "exists" clause will be printed.
+
S+poonceonces.litmus
As below, but without the smp_wmb() and acquire load.
diff --git a/tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus b/tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus
new file mode 100644
index 000000000000..04a16603660b
--- /dev/null
+++ b/tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus
@@ -0,0 +1,32 @@
+C SB+rfionceonce-poonceonces
+
+(*
+ * Result: Sometimes
+ *
+ * This litmus test demonstrates that LKMM is not fully multicopy atomic.
+ *)
+
+{}
+
+P0(int *x, int *y)
+{
+ int r1;
+ int r2;
+
+ WRITE_ONCE(*x, 1);
+ r1 = READ_ONCE(*x);
+ r2 = READ_ONCE(*y);
+}
+
+P1(int *x, int *y)
+{
+ int r3;
+ int r4;
+
+ WRITE_ONCE(*y, 1);
+ r3 = READ_ONCE(*y);
+ r4 = READ_ONCE(*x);
+}
+
+locations [0:r1; 1:r3; x; y] (* Debug aid: Print things not in "exists". *)
+exists (0:r2=0 /\ 1:r4=0)
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 02/14] tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 01/14] tools/memory-model: Add litmus test for full multicopy atomicity Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:26 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 03/14] MAINTAINERS: Add Daniel Lustig as an LKMM reviewer Paul E. McKenney
` (11 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Paul E. McKenney
The names on the first line of the litmus tests are arbitrary,
but the convention is that they be the filename without the trailing
".litmus". This commit therefore removes the stray trailing ".litmus"
from ISA2+pooncelock+pooncelock+pombonce.litmus's name.
Reported-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
---
.../litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus b/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus
index 7a39a0aaa976..0f749e419b34 100644
--- a/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus
+++ b/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus
@@ -1,4 +1,4 @@
-C ISA2+pooncelock+pooncelock+pombonce.litmus
+C ISA2+pooncelock+pooncelock+pombonce
(*
* Result: Sometimes
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 03/14] MAINTAINERS: Add Daniel Lustig as an LKMM reviewer
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 01/14] tools/memory-model: Add litmus test for full multicopy atomicity Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 02/14] tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:26 ` [tip:locking/core] " tip-bot for Palmer Dabbelt
2018-07-16 18:05 ` [PATCH memory-model 04/14] locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example Paul E. McKenney
` (10 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Palmer Dabbelt,
Daniel Lustig, Paul E . McKenney
From: Palmer Dabbelt <palmer@sifive.com>
Dan runs the RISC-V memory model working group. I've been forwarding
him LKMM emails that end up in my inbox, but I'm far from an expert in
this stuff. He requested to be added as a reviewer, which seems sane to
me as it'll take a human out of the loop.
CC: Daniel Lustig <dlustig@nvidia.com>
Acked-by: Daniel Lustig <dlustig@nvidia.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Palmer Dabbelt <palmer@sifive.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 9d5eeff51b5f..ac8ed55dbe9b 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8316,6 +8316,7 @@ M: Jade Alglave <j.alglave@ucl.ac.uk>
M: Luc Maranget <luc.maranget@inria.fr>
M: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
R: Akira Yokosawa <akiyks@gmail.com>
+R: Daniel Lustig <dlustig@nvidia.com>
L: linux-kernel@vger.kernel.org
S: Supported
T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 04/14] locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (2 preceding siblings ...)
2018-07-16 18:05 ` [PATCH memory-model 03/14] MAINTAINERS: Add Daniel Lustig as an LKMM reviewer Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:27 ` [tip:locking/core] " tip-bot for SeongJae Park
2018-07-16 18:05 ` [PATCH memory-model 05/14] tools/memory-model: Remove ACCESS_ONCE() from recipes Paul E. McKenney
` (9 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, SeongJae Park,
Paul E . McKenney
From: SeongJae Park <sj38.park@gmail.com>
Translate this commit to Korean:
5846581e3563 ("locking/memory-barriers.txt: Fix broken DMA vs. MMIO ordering example")
Signed-off-by: SeongJae Park <sj38.park@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
[ paulmck: Updated based on feedback from Byungchul Park. ]
Acked-by: Byungchul Park <byungchul.park@lge.com>
---
.../translations/ko_KR/memory-barriers.txt | 22 +++++++++----------
1 file changed, 11 insertions(+), 11 deletions(-)
diff --git a/Documentation/translations/ko_KR/memory-barriers.txt b/Documentation/translations/ko_KR/memory-barriers.txt
index 921739d00f69..7f01fb1c1084 100644
--- a/Documentation/translations/ko_KR/memory-barriers.txt
+++ b/Documentation/translations/ko_KR/memory-barriers.txt
@@ -1891,22 +1891,22 @@ Mandatory 배리어들은 SMP 시스템에서도 UP 시스템에서도 SMP 효
/* 소유권을 수정 */
desc->status = DEVICE_OWN;
- /* MMIO 를 통해 디바이스에 공지를 하기 전에 메모리를 동기화 */
- wmb();
-
/* 업데이트된 디스크립터의 디바이스에 공지 */
writel(DESC_NOTIFY, doorbell);
}
dma_rmb() 는 디스크립터로부터 데이터를 읽어오기 전에 디바이스가 소유권을
- 내놓았음을 보장하게 하고, dma_wmb() 는 디바이스가 자신이 소유권을 다시
- 가졌음을 보기 전에 디스크립터에 데이터가 쓰였음을 보장합니다. wmb() 는
- 캐시 일관성이 없는 (cache incoherent) MMIO 영역에 쓰기를 시도하기 전에
- 캐시 일관성이 있는 메모리 (cache coherent memory) 쓰기가 완료되었음을
- 보장해주기 위해 필요합니다.
-
- consistent memory 에 대한 자세한 내용을 위해선 Documentation/DMA-API.txt
- 문서를 참고하세요.
+ 내려놓았을 것을 보장하고, dma_wmb() 는 디바이스가 자신이 소유권을 다시
+ 가졌음을 보기 전에 디스크립터에 데이터가 쓰였을 것을 보장합니다. 참고로,
+ writel() 을 사용하면 캐시 일관성이 있는 메모리 (cache coherent memory)
+ 쓰기가 MMIO 영역에의 쓰기 전에 완료되었을 것을 보장하므로 writel() 앞에
+ wmb() 를 실행할 필요가 없음을 알아두시기 바랍니다. writel() 보다 비용이
+ 저렴한 writel_relaxed() 는 이런 보장을 제공하지 않으므로 여기선 사용되지
+ 않아야 합니다.
+
+ writel_relaxed() 와 같은 완화된 I/O 접근자들에 대한 자세한 내용을 위해서는
+ "커널 I/O 배리어의 효과" 섹션을, consistent memory 에 대한 자세한 내용을
+ 위해선 Documentation/DMA-API.txt 문서를 참고하세요.
MMIO 쓰기 배리어
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 05/14] tools/memory-model: Remove ACCESS_ONCE() from recipes
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (3 preceding siblings ...)
2018-07-16 18:05 ` [PATCH memory-model 04/14] locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:27 ` [tip:locking/core] " tip-bot for Mark Rutland
2018-07-16 18:05 ` [PATCH memory-model 06/14] tools/memory-model: Remove ACCESS_ONCE() from model Paul E. McKenney
` (8 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Mark Rutland,
Paul E . McKenney
From: Mark Rutland <mark.rutland@arm.com>
Since commit:
b899a850431e2dd0 ("compiler.h: Remove ACCESS_ONCE()")
... there has been no definition of ACCESS_ONCE() in the kernel tree,
and it has been necessary to use READ_ONCE() or WRITE_ONCE() instead.
Let's update the exmaples in recipes.txt likewise for consistency, using
READ_ONCE() for reads.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Akira Yokosawa <akiyks@gmail.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
tools/memory-model/Documentation/recipes.txt | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/memory-model/Documentation/recipes.txt b/tools/memory-model/Documentation/recipes.txt
index ee4309a87fc4..1fea8ef2b184 100644
--- a/tools/memory-model/Documentation/recipes.txt
+++ b/tools/memory-model/Documentation/recipes.txt
@@ -322,9 +322,9 @@ the following write-side code fragment:
And the xlog_valid_lsn() function in fs/xfs/xfs_log_priv.h contains
the corresponding read-side code fragment:
- cur_cycle = ACCESS_ONCE(log->l_curr_cycle);
+ cur_cycle = READ_ONCE(log->l_curr_cycle);
smp_rmb();
- cur_block = ACCESS_ONCE(log->l_curr_block);
+ cur_block = READ_ONCE(log->l_curr_block);
Alternatively, consider the following comment in function
perf_output_put_handle() in kernel/events/ring_buffer.c:
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 06/14] tools/memory-model: Remove ACCESS_ONCE() from model
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (4 preceding siblings ...)
2018-07-16 18:05 ` [PATCH memory-model 05/14] tools/memory-model: Remove ACCESS_ONCE() from recipes Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:28 ` [tip:locking/core] " tip-bot for Mark Rutland
2018-07-16 18:05 ` [PATCH memory-model 07/14] tools/memory-model: Make scripts executable Paul E. McKenney
` (7 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Mark Rutland,
Paul E . McKenney
From: Mark Rutland <mark.rutland@arm.com>
Since commit:
b899a850431e2dd0 ("compiler.h: Remove ACCESS_ONCE()")
... there has been no definition of ACCESS_ONCE() in the kernel tree,
and it has been necessary to use READ_ONCE() or WRITE_ONCE() instead.
Correspondingly, let's remove ACCESS_ONCE() from the kernel memory
model.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Akira Yokosawa <akiyks@gmail.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
tools/memory-model/linux-kernel.bell | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/memory-model/linux-kernel.bell b/tools/memory-model/linux-kernel.bell
index 64f5740e0e75..b84fb2f67109 100644
--- a/tools/memory-model/linux-kernel.bell
+++ b/tools/memory-model/linux-kernel.bell
@@ -13,7 +13,7 @@
"Linux-kernel memory consistency model"
-enum Accesses = 'once (*READ_ONCE,WRITE_ONCE,ACCESS_ONCE*) ||
+enum Accesses = 'once (*READ_ONCE,WRITE_ONCE*) ||
'release (*smp_store_release*) ||
'acquire (*smp_load_acquire*) ||
'noreturn (* R of non-return RMW *)
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 07/14] tools/memory-model: Make scripts executable
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (5 preceding siblings ...)
2018-07-16 18:05 ` [PATCH memory-model 06/14] tools/memory-model: Remove ACCESS_ONCE() from model Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-17 9:29 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 08/14] docs: atomic_ops: Describe atomic_set as a write operation Paul E. McKenney
` (6 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Paul E. McKenney
This commit makes the scripts executable to avoid the need for everyone
to do so manually in their archive.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Akira Yokosawa <akiyks@gmail.com>
---
tools/memory-model/scripts/checkalllitmus.sh | 2 +-
tools/memory-model/scripts/checklitmus.sh | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
mode change 100644 => 100755 tools/memory-model/scripts/checkalllitmus.sh
mode change 100644 => 100755 tools/memory-model/scripts/checklitmus.sh
diff --git a/tools/memory-model/scripts/checkalllitmus.sh b/tools/memory-model/scripts/checkalllitmus.sh
old mode 100644
new mode 100755
index af0aa15ab84e..ca528f9a24d4
--- a/tools/memory-model/scripts/checkalllitmus.sh
+++ b/tools/memory-model/scripts/checkalllitmus.sh
@@ -9,7 +9,7 @@
# appended.
#
# Usage:
-# sh checkalllitmus.sh [ directory ]
+# checkalllitmus.sh [ directory ]
#
# The LINUX_HERD_OPTIONS environment variable may be used to specify
# arguments to herd, whose default is defined by the checklitmus.sh script.
diff --git a/tools/memory-model/scripts/checklitmus.sh b/tools/memory-model/scripts/checklitmus.sh
old mode 100644
new mode 100755
index e2e477472844..bf12a75c0719
--- a/tools/memory-model/scripts/checklitmus.sh
+++ b/tools/memory-model/scripts/checklitmus.sh
@@ -8,7 +8,7 @@
# with ".out" appended.
#
# Usage:
-# sh checklitmus.sh file.litmus
+# checklitmus.sh file.litmus
#
# The LINUX_HERD_OPTIONS environment variable may be used to specify
# arguments to herd, which default to "-conf linux-kernel.cfg". Thus,
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 08/14] docs: atomic_ops: Describe atomic_set as a write operation
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (6 preceding siblings ...)
2018-07-16 18:05 ` [PATCH memory-model 07/14] tools/memory-model: Make scripts executable Paul E. McKenney
@ 2018-07-16 18:05 ` Paul E. McKenney
2018-07-16 18:06 ` [PATCH memory-model 09/14] tools/memory-model: Add informal LKMM documentation to MAINTAINERS Paul E. McKenney
` (5 subsequent siblings)
13 siblings, 0 replies; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:05 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks,
Jonathan Neuschäfer, Paul E . McKenney
From: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
The atomic_set() and ATOMIC_INIT() operations are writes, so this
commit changes their description from "reads" to "writes".
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Andrea Parri <andrea.parri@amarulasolutions.com>
---
Documentation/core-api/atomic_ops.rst | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Documentation/core-api/atomic_ops.rst b/Documentation/core-api/atomic_ops.rst
index 2e7165f86f55..724583453e1f 100644
--- a/Documentation/core-api/atomic_ops.rst
+++ b/Documentation/core-api/atomic_ops.rst
@@ -29,7 +29,7 @@ updated by one CPU, local_t is probably more appropriate. Please see
local_t.
The first operations to implement for atomic_t's are the initializers and
-plain reads. ::
+plain writes. ::
#define ATOMIC_INIT(i) { (i) }
#define atomic_set(v, i) ((v)->counter = (i))
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 09/14] tools/memory-model: Add informal LKMM documentation to MAINTAINERS
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (7 preceding siblings ...)
2018-07-16 18:05 ` [PATCH memory-model 08/14] docs: atomic_ops: Describe atomic_set as a write operation Paul E. McKenney
@ 2018-07-16 18:06 ` Paul E. McKenney
2018-07-17 9:30 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:06 ` [PATCH memory-model 10/14] sched: Use smp_mb() in wake_woken_function() Paul E. McKenney
` (4 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:06 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Paul E. McKenney,
Daniel Lustig, David S. Miller
The Linux-kernel memory model has been informal, with a number of
text files documenting it. It would be good to make sure that these
informal descriptions are kept up to date and/or pruned appropriately.
This commit therefore brings more of those text files into the LKMM
MAINTAINERS file entry.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Daniel Lustig <dlustig@nvidia.com>
Cc: "David S. Miller" <davem@davemloft.net>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
---
MAINTAINERS | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index ac8ed55dbe9b..b60d6daad132 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8318,9 +8318,14 @@ M: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
R: Akira Yokosawa <akiyks@gmail.com>
R: Daniel Lustig <dlustig@nvidia.com>
L: linux-kernel@vger.kernel.org
+L: linux-arch@vger.kernel.org
S: Supported
T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
F: tools/memory-model/
+F: Documentation/atomic_bitops.txt
+F: Documentation/atomic_t.txt
+F: Documentation/core-api/atomic_ops.rst
+F: Documentation/core-api/refcount-vs-atomic.rst
F: Documentation/memory-barriers.txt
LINUX SECURITY MODULE (LSM) FRAMEWORK
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 10/14] sched: Use smp_mb() in wake_woken_function()
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (8 preceding siblings ...)
2018-07-16 18:06 ` [PATCH memory-model 09/14] tools/memory-model: Add informal LKMM documentation to MAINTAINERS Paul E. McKenney
@ 2018-07-16 18:06 ` Paul E. McKenney
2018-07-17 9:30 ` [tip:locking/core] sched/core: " tip-bot for Andrea Parri
2018-07-16 18:06 ` [PATCH memory-model 11/14] locking: Clarify requirements for smp_mb__after_spinlock() Paul E. McKenney
` (3 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:06 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Andrea Parri,
Ingo Molnar, Paul E . McKenney
From: Andrea Parri <andrea.parri@amarulasolutions.com>
wake_woken_function() synchronizes with wait_woken() as follows:
[wait_woken] [wake_woken_function]
entry->flags &= ~wq_flag_woken; condition = true;
smp_mb(); smp_wmb();
if (condition) wq_entry->flags |= wq_flag_woken;
break;
This commit replaces the above smp_wmb() with an smp_mb() in order to
guarantee that either wait_woken() sees the wait condition being true
or the store to wq_entry->flags in woken_wake_function() follows the
store in wait_woken() in the coherence order (so that the former can
eventually be observed by wait_woken()).
The commit also fixes a comment associated to set_current_state() in
wait_woken(): the comment pairs the barrier in set_current_state() to
the above smp_wmb(), while the actual pairing involves the barrier in
set_current_state() and the barrier executed by the try_to_wake_up()
in wake_woken_function().
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Ingo Molnar <mingo@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
kernel/sched/wait.c | 47 ++++++++++++++++++++-------------------------
1 file changed, 21 insertions(+), 26 deletions(-)
diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c
index 928be527477e..a7a2aaa3026a 100644
--- a/kernel/sched/wait.c
+++ b/kernel/sched/wait.c
@@ -392,35 +392,36 @@ static inline bool is_kthread_should_stop(void)
* if (condition)
* break;
*
- * p->state = mode; condition = true;
- * smp_mb(); // A smp_wmb(); // C
- * if (!wq_entry->flags & WQ_FLAG_WOKEN) wq_entry->flags |= WQ_FLAG_WOKEN;
- * schedule() try_to_wake_up();
- * p->state = TASK_RUNNING; ~~~~~~~~~~~~~~~~~~
- * wq_entry->flags &= ~WQ_FLAG_WOKEN; condition = true;
- * smp_mb() // B smp_wmb(); // C
- * wq_entry->flags |= WQ_FLAG_WOKEN;
- * }
- * remove_wait_queue(&wq_head, &wait);
+ * // in wait_woken() // in woken_wake_function()
*
+ * p->state = mode; wq_entry->flags |= WQ_FLAG_WOKEN;
+ * smp_mb(); // A try_to_wake_up():
+ * if (!(wq_entry->flags & WQ_FLAG_WOKEN)) <full barrier>
+ * schedule() if (p->state & mode)
+ * p->state = TASK_RUNNING; p->state = TASK_RUNNING;
+ * wq_entry->flags &= ~WQ_FLAG_WOKEN; ~~~~~~~~~~~~~~~~~~
+ * smp_mb(); // B condition = true;
+ * } smp_mb(); // C
+ * remove_wait_queue(&wq_head, &wait); wq_entry->flags |= WQ_FLAG_WOKEN;
*/
long wait_woken(struct wait_queue_entry *wq_entry, unsigned mode, long timeout)
{
- set_current_state(mode); /* A */
/*
- * The above implies an smp_mb(), which matches with the smp_wmb() from
- * woken_wake_function() such that if we observe WQ_FLAG_WOKEN we must
- * also observe all state before the wakeup.
+ * The below executes an smp_mb(), which matches with the full barrier
+ * executed by the try_to_wake_up() in woken_wake_function() such that
+ * either we see the store to wq_entry->flags in woken_wake_function()
+ * or woken_wake_function() sees our store to current->state.
*/
+ set_current_state(mode); /* A */
if (!(wq_entry->flags & WQ_FLAG_WOKEN) && !is_kthread_should_stop())
timeout = schedule_timeout(timeout);
__set_current_state(TASK_RUNNING);
/*
- * The below implies an smp_mb(), it too pairs with the smp_wmb() from
- * woken_wake_function() such that we must either observe the wait
- * condition being true _OR_ WQ_FLAG_WOKEN such that we will not miss
- * an event.
+ * The below executes an smp_mb(), which matches with the smp_mb() (C)
+ * in woken_wake_function() such that either we see the wait condition
+ * being true or the store to wq_entry->flags in woken_wake_function()
+ * follows ours in the coherence order.
*/
smp_store_mb(wq_entry->flags, wq_entry->flags & ~WQ_FLAG_WOKEN); /* B */
@@ -430,14 +431,8 @@ EXPORT_SYMBOL(wait_woken);
int woken_wake_function(struct wait_queue_entry *wq_entry, unsigned mode, int sync, void *key)
{
- /*
- * Although this function is called under waitqueue lock, LOCK
- * doesn't imply write barrier and the users expects write
- * barrier semantics on wakeup functions. The following
- * smp_wmb() is equivalent to smp_wmb() in try_to_wake_up()
- * and is paired with smp_store_mb() in wait_woken().
- */
- smp_wmb(); /* C */
+ /* Pairs with the smp_store_mb() in wait_woken(). */
+ smp_mb(); /* C */
wq_entry->flags |= WQ_FLAG_WOKEN;
return default_wake_function(wq_entry, mode, sync, key);
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 11/14] locking: Clarify requirements for smp_mb__after_spinlock()
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (9 preceding siblings ...)
2018-07-16 18:06 ` [PATCH memory-model 10/14] sched: Use smp_mb() in wake_woken_function() Paul E. McKenney
@ 2018-07-16 18:06 ` Paul E. McKenney
2018-07-17 9:31 ` [tip:locking/core] locking/spinlock, sched/core: " tip-bot for Andrea Parri
2018-07-16 18:06 ` [PATCH memory-model 12/14] doc: Update wake_up() & co. memory-barrier guarantees Paul E. McKenney
` (2 subsequent siblings)
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:06 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Andrea Parri,
Ingo Molnar, Paul E. McKenney
From: Andrea Parri <andrea.parri@amarulasolutions.com>
There are 11 interpretations of the requirements described in the header
comment for smp_mb__after_spinlock(): one for each LKMM maintainer, and
one currently encoded in the Cat file. Stick to the latter (until a more
satisfactory solution is available).
This also reworks some snippets related to the barrier to illustrate the
requirements and to link them to the idioms which are relied upon at its
call sites.
Suggested-by: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
include/linux/spinlock.h | 53 +++++++++++++++++++++++++++-------------
kernel/sched/core.c | 41 ++++++++++++++++---------------
2 files changed, 57 insertions(+), 37 deletions(-)
diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
index 1e8a46435838..d70a06ff2bdd 100644
--- a/include/linux/spinlock.h
+++ b/include/linux/spinlock.h
@@ -114,29 +114,48 @@ do { \
#endif /*arch_spin_is_contended*/
/*
- * This barrier must provide two things:
+ * smp_mb__after_spinlock() provides the equivalent of a full memory barrier
+ * between program-order earlier lock acquisitions and program-order later
+ * memory accesses.
*
- * - it must guarantee a STORE before the spin_lock() is ordered against a
- * LOAD after it, see the comments at its two usage sites.
+ * This guarantees that the following two properties hold:
*
- * - it must ensure the critical section is RCsc.
+ * 1) Given the snippet:
*
- * The latter is important for cases where we observe values written by other
- * CPUs in spin-loops, without barriers, while being subject to scheduling.
+ * { X = 0; Y = 0; }
*
- * CPU0 CPU1 CPU2
+ * CPU0 CPU1
*
- * for (;;) {
- * if (READ_ONCE(X))
- * break;
- * }
- * X=1
- * <sched-out>
- * <sched-in>
- * r = X;
+ * WRITE_ONCE(X, 1); WRITE_ONCE(Y, 1);
+ * spin_lock(S); smp_mb();
+ * smp_mb__after_spinlock(); r1 = READ_ONCE(X);
+ * r0 = READ_ONCE(Y);
+ * spin_unlock(S);
*
- * without transitivity it could be that CPU1 observes X!=0 breaks the loop,
- * we get migrated and CPU2 sees X==0.
+ * it is forbidden that CPU0 does not observe CPU1's store to Y (r0 = 0)
+ * and CPU1 does not observe CPU0's store to X (r1 = 0); see the comments
+ * preceding the call to smp_mb__after_spinlock() in __schedule() and in
+ * try_to_wake_up().
+ *
+ * 2) Given the snippet:
+ *
+ * { X = 0; Y = 0; }
+ *
+ * CPU0 CPU1 CPU2
+ *
+ * spin_lock(S); spin_lock(S); r1 = READ_ONCE(Y);
+ * WRITE_ONCE(X, 1); smp_mb__after_spinlock(); smp_rmb();
+ * spin_unlock(S); r0 = READ_ONCE(X); r2 = READ_ONCE(X);
+ * WRITE_ONCE(Y, 1);
+ * spin_unlock(S);
+ *
+ * it is forbidden that CPU0's critical section executes before CPU1's
+ * critical section (r0 = 1), CPU2 observes CPU1's store to Y (r1 = 1)
+ * and CPU2 does not observe CPU0's store to X (r2 = 0); see the comments
+ * preceding the calls to smp_rmb() in try_to_wake_up() for similar
+ * snippets but "projected" onto two CPUs.
+ *
+ * Property (2) upgrades the lock to an RCsc lock.
*
* Since most load-store architectures implement ACQUIRE with an smp_mb() after
* the LL/SC loop, they need no further barriers. Similarly all our TSO
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 78d8facba456..7db0662360f1 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1999,21 +1999,20 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
* be possible to, falsely, observe p->on_rq == 0 and get stuck
* in smp_cond_load_acquire() below.
*
- * sched_ttwu_pending() try_to_wake_up()
- * [S] p->on_rq = 1; [L] P->state
- * UNLOCK rq->lock -----.
- * \
- * +--- RMB
- * schedule() /
- * LOCK rq->lock -----'
- * UNLOCK rq->lock
+ * sched_ttwu_pending() try_to_wake_up()
+ * STORE p->on_rq = 1 LOAD p->state
+ * UNLOCK rq->lock
+ *
+ * __schedule() (switch to task 'p')
+ * LOCK rq->lock smp_rmb();
+ * smp_mb__after_spinlock();
+ * UNLOCK rq->lock
*
* [task p]
- * [S] p->state = UNINTERRUPTIBLE [L] p->on_rq
+ * STORE p->state = UNINTERRUPTIBLE LOAD p->on_rq
*
- * Pairs with the UNLOCK+LOCK on rq->lock from the
- * last wakeup of our task and the schedule that got our task
- * current.
+ * Pairs with the LOCK+smp_mb__after_spinlock() on rq->lock in
+ * __schedule(). See the comment for smp_mb__after_spinlock().
*/
smp_rmb();
if (p->on_rq && ttwu_remote(p, wake_flags))
@@ -2027,15 +2026,17 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
* One must be running (->on_cpu == 1) in order to remove oneself
* from the runqueue.
*
- * [S] ->on_cpu = 1; [L] ->on_rq
- * UNLOCK rq->lock
- * RMB
- * LOCK rq->lock
- * [S] ->on_rq = 0; [L] ->on_cpu
+ * __schedule() (switch to task 'p') try_to_wake_up()
+ * STORE p->on_cpu = 1 LOAD p->on_rq
+ * UNLOCK rq->lock
+ *
+ * __schedule() (put 'p' to sleep)
+ * LOCK rq->lock smp_rmb();
+ * smp_mb__after_spinlock();
+ * STORE p->on_rq = 0 LOAD p->on_cpu
*
- * Pairs with the full barrier implied in the UNLOCK+LOCK on rq->lock
- * from the consecutive calls to schedule(); the first switching to our
- * task, the second putting it to sleep.
+ * Pairs with the LOCK+smp_mb__after_spinlock() on rq->lock in
+ * __schedule(). See the comment for smp_mb__after_spinlock().
*/
smp_rmb();
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 12/14] doc: Update wake_up() & co. memory-barrier guarantees
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (10 preceding siblings ...)
2018-07-16 18:06 ` [PATCH memory-model 11/14] locking: Clarify requirements for smp_mb__after_spinlock() Paul E. McKenney
@ 2018-07-16 18:06 ` Paul E. McKenney
2018-07-17 9:31 ` [tip:locking/core] sched/Documentation: " tip-bot for Andrea Parri
2018-07-16 18:06 ` [PATCH memory-model 13/14] memory-model/Documentation: Fix typo, smb->smp Paul E. McKenney
2018-07-16 18:06 ` [PATCH memory-model 14/14] tools/memory-model: Rename litmus tests to comply to norm7 Paul E. McKenney
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:06 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Andrea Parri,
Daniel Lustig, Jonathan Corbet, Ingo Molnar, Paul E . McKenney
From: Andrea Parri <andrea.parri@amarulasolutions.com>
Both the implementation and the users' expectation [1] for the various
wakeup primitives have evolved over time, but the documentation has not
kept up with these changes: brings it into 2018.
[1] http://lkml.kernel.org/r/20180424091510.GB4064@hirez.programming.kicks-ass.net
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
[ aparri: Apply feedback from Alan Stern. ]
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Daniel Lustig <dlustig@nvidia.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Ingo Molnar <mingo@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
Documentation/memory-barriers.txt | 43 +++++++++++++++++++------------
include/linux/sched.h | 4 +--
kernel/sched/completion.c | 8 +++---
kernel/sched/core.c | 30 +++++++++------------
kernel/sched/wait.c | 8 +++---
5 files changed, 49 insertions(+), 44 deletions(-)
diff --git a/Documentation/memory-barriers.txt b/Documentation/memory-barriers.txt
index a02d6bbfc9d0..0d8d7ef131e9 100644
--- a/Documentation/memory-barriers.txt
+++ b/Documentation/memory-barriers.txt
@@ -2179,32 +2179,41 @@ or:
event_indicated = 1;
wake_up_process(event_daemon);
-A write memory barrier is implied by wake_up() and co. if and only if they
-wake something up. The barrier occurs before the task state is cleared, and so
-sits between the STORE to indicate the event and the STORE to set TASK_RUNNING:
+A general memory barrier is executed by wake_up() if it wakes something up.
+If it doesn't wake anything up then a memory barrier may or may not be
+executed; you must not rely on it. The barrier occurs before the task state
+is accessed, in particular, it sits between the STORE to indicate the event
+and the STORE to set TASK_RUNNING:
- CPU 1 CPU 2
+ CPU 1 (Sleeper) CPU 2 (Waker)
=============================== ===============================
set_current_state(); STORE event_indicated
smp_store_mb(); wake_up();
- STORE current->state <write barrier>
- <general barrier> STORE current->state
- LOAD event_indicated
+ STORE current->state ...
+ <general barrier> <general barrier>
+ LOAD event_indicated if ((LOAD task->state) & TASK_NORMAL)
+ STORE task->state
-To repeat, this write memory barrier is present if and only if something
-is actually awakened. To see this, consider the following sequence of
-events, where X and Y are both initially zero:
+where "task" is the thread being woken up and it equals CPU 1's "current".
+
+To repeat, a general memory barrier is guaranteed to be executed by wake_up()
+if something is actually awakened, but otherwise there is no such guarantee.
+To see this, consider the following sequence of events, where X and Y are both
+initially zero:
CPU 1 CPU 2
=============================== ===============================
- X = 1; STORE event_indicated
+ X = 1; Y = 1;
smp_mb(); wake_up();
- Y = 1; wait_event(wq, Y == 1);
- wake_up(); load from Y sees 1, no memory barrier
- load from X might see 0
+ LOAD Y LOAD X
+
+If a wakeup does occur, one (at least) of the two loads must see 1. If, on
+the other hand, a wakeup does not occur, both loads might see 0.
-In contrast, if a wakeup does occur, CPU 2's load from X would be guaranteed
-to see 1.
+wake_up_process() always executes a general memory barrier. The barrier again
+occurs before the task state is accessed. In particular, if the wake_up() in
+the previous snippet were replaced by a call to wake_up_process() then one of
+the two loads would be guaranteed to see 1.
The available waker functions include:
@@ -2224,6 +2233,8 @@ The available waker functions include:
wake_up_poll();
wake_up_process();
+In terms of memory ordering, these functions all provide the same guarantees of
+a wake_up() (or stronger).
[!] Note that the memory barriers implied by the sleeper and the waker do _not_
order multiple stores before the wake-up with respect to loads of those stored
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 87bf02d93a27..ddfdeb632f74 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -167,8 +167,8 @@ struct task_group;
* need_sleep = false;
* wake_up_state(p, TASK_UNINTERRUPTIBLE);
*
- * Where wake_up_state() (and all other wakeup primitives) imply enough
- * barriers to order the store of the variable against wakeup.
+ * where wake_up_state() executes a full memory barrier before accessing the
+ * task state.
*
* Wakeup will do: if (@state & p->state) p->state = TASK_RUNNING, that is,
* once it observes the TASK_UNINTERRUPTIBLE store the waking CPU can issue a
diff --git a/kernel/sched/completion.c b/kernel/sched/completion.c
index e426b0cb9ac6..a1ad5b7d5521 100644
--- a/kernel/sched/completion.c
+++ b/kernel/sched/completion.c
@@ -22,8 +22,8 @@
*
* See also complete_all(), wait_for_completion() and related routines.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*/
void complete(struct completion *x)
{
@@ -44,8 +44,8 @@ EXPORT_SYMBOL(complete);
*
* This will wake up all threads waiting on this particular completion event.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*
* Since complete_all() sets the completion of @x permanently to done
* to allow multiple waiters to finish, a call to reinit_completion()
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 7db0662360f1..940dfdae41b5 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -413,8 +413,8 @@ void wake_q_add(struct wake_q_head *head, struct task_struct *task)
* its already queued (either by us or someone else) and will get the
* wakeup due to that.
*
- * This cmpxchg() implies a full barrier, which pairs with the write
- * barrier implied by the wakeup in wake_up_q().
+ * This cmpxchg() executes a full barrier, which pairs with the full
+ * barrier executed by the wakeup in wake_up_q().
*/
if (cmpxchg(&node->next, NULL, WAKE_Q_TAIL))
return;
@@ -442,8 +442,8 @@ void wake_up_q(struct wake_q_head *head)
task->wake_q.next = NULL;
/*
- * wake_up_process() implies a wmb() to pair with the queueing
- * in wake_q_add() so as not to miss wakeups.
+ * wake_up_process() executes a full barrier, which pairs with
+ * the queueing in wake_q_add() so as not to miss wakeups.
*/
wake_up_process(task);
put_task_struct(task);
@@ -1880,8 +1880,7 @@ static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags)
* rq(c1)->lock (if not at the same time, then in that order).
* C) LOCK of the rq(c1)->lock scheduling in task
*
- * Transitivity guarantees that B happens after A and C after B.
- * Note: we only require RCpc transitivity.
+ * Release/acquire chaining guarantees that B happens after A and C after B.
* Note: the CPU doing B need not be c0 or c1
*
* Example:
@@ -1943,16 +1942,9 @@ static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags)
* UNLOCK rq(0)->lock
*
*
- * However; for wakeups there is a second guarantee we must provide, namely we
- * must observe the state that lead to our wakeup. That is, not only must our
- * task observe its own prior state, it must also observe the stores prior to
- * its wakeup.
- *
- * This means that any means of doing remote wakeups must order the CPU doing
- * the wakeup against the CPU the task is going to end up running on. This,
- * however, is already required for the regular Program-Order guarantee above,
- * since the waking CPU is the one issueing the ACQUIRE (smp_cond_load_acquire).
- *
+ * However, for wakeups there is a second guarantee we must provide, namely we
+ * must ensure that CONDITION=1 done by the caller can not be reordered with
+ * accesses to the task state; see try_to_wake_up() and set_current_state().
*/
/**
@@ -1968,6 +1960,9 @@ static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags)
* Atomic against schedule() which would dequeue a task, also see
* set_current_state().
*
+ * This function executes a full memory barrier before accessing the task
+ * state; see set_current_state().
+ *
* Return: %true if @p->state changes (an actual wakeup was done),
* %false otherwise.
*/
@@ -2142,8 +2137,7 @@ static void try_to_wake_up_local(struct task_struct *p, struct rq_flags *rf)
*
* Return: 1 if the process was woken up, 0 if it was already running.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * This function executes a full memory barrier before accessing the task state.
*/
int wake_up_process(struct task_struct *p)
{
diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c
index a7a2aaa3026a..870f97b313e3 100644
--- a/kernel/sched/wait.c
+++ b/kernel/sched/wait.c
@@ -134,8 +134,8 @@ static void __wake_up_common_lock(struct wait_queue_head *wq_head, unsigned int
* @nr_exclusive: how many wake-one or wake-many threads to wake up
* @key: is directly passed to the wakeup function
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*/
void __wake_up(struct wait_queue_head *wq_head, unsigned int mode,
int nr_exclusive, void *key)
@@ -180,8 +180,8 @@ EXPORT_SYMBOL_GPL(__wake_up_locked_key_bookmark);
*
* On UP it can prevent extra preemption.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*/
void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode,
int nr_exclusive, void *key)
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 13/14] memory-model/Documentation: Fix typo, smb->smp
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (11 preceding siblings ...)
2018-07-16 18:06 ` [PATCH memory-model 12/14] doc: Update wake_up() & co. memory-barrier guarantees Paul E. McKenney
@ 2018-07-16 18:06 ` Paul E. McKenney
2018-07-17 9:32 ` [tip:locking/core] tools/memory-model/Documentation: " tip-bot for Yauheni Kaliuta
2018-07-16 18:06 ` [PATCH memory-model 14/14] tools/memory-model: Rename litmus tests to comply to norm7 Paul E. McKenney
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:06 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Yauheni Kaliuta,
Paul E . McKenney
From: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
The tools/memory-model/Documentation/explanation.txt file says
"For each other CPU C', smb_wmb() forces all po-earlier stores"
This commit therefore replaces the "smb_wmb()" with "smp_wmb()".
Signed-off-by: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
---
tools/memory-model/Documentation/explanation.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/memory-model/Documentation/explanation.txt b/tools/memory-model/Documentation/explanation.txt
index 1b09f3175a1f..0cbd1ef8f86d 100644
--- a/tools/memory-model/Documentation/explanation.txt
+++ b/tools/memory-model/Documentation/explanation.txt
@@ -804,7 +804,7 @@ type of fence:
Second, some types of fence affect the way the memory subsystem
propagates stores. When a fence instruction is executed on CPU C:
- For each other CPU C', smb_wmb() forces all po-earlier stores
+ For each other CPU C', smp_wmb() forces all po-earlier stores
on C to propagate to C' before any po-later stores do.
For each other CPU C', any store which propagates to C before
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [PATCH memory-model 14/14] tools/memory-model: Rename litmus tests to comply to norm7
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
` (12 preceding siblings ...)
2018-07-16 18:06 ` [PATCH memory-model 13/14] memory-model/Documentation: Fix typo, smb->smp Paul E. McKenney
@ 2018-07-16 18:06 ` Paul E. McKenney
2018-07-17 9:33 ` [tip:locking/core] " tip-bot for Andrea Parri
13 siblings, 1 reply; 28+ messages in thread
From: Paul E. McKenney @ 2018-07-16 18:06 UTC (permalink / raw)
To: linux-kernel, linux-arch, mingo
Cc: stern, parri.andrea, will.deacon, peterz, boqun.feng, npiggin,
dhowells, j.alglave, luc.maranget, akiyks, Andrea Parri,
Paul E. McKenney
From: Andrea Parri <andrea.parri@amarulasolutions.com>
norm7 produces the 'normalized' name of a litmus test, when the test
can be generated from a single cycle that passes through each process
exactly once. The commit renames such tests in order to comply to the
naming scheme implemented by this tool.
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Akira Yokosawa <akiyks@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
---
tools/memory-model/Documentation/recipes.txt | 8 ++++----
tools/memory-model/README | 20 +++++++++----------
... => IRIW+fencembonceonces+OnceOnce.litmus} | 2 +-
...=> LB+fencembonceonce+ctrlonceonce.litmus} | 2 +-
...+fencewmbonceonce+fencermbonceonce.litmus} | 2 +-
...onces.litmus => R+fencembonceonces.litmus} | 2 +-
tools/memory-model/litmus-tests/README | 16 +++++++--------
...> S+fencewmbonceonce+poacquireonce.litmus} | 2 +-
...nces.litmus => SB+fencembonceonces.litmus} | 2 +-
...ooncerelease+fencermbonceonce+Once.litmus} | 2 +-
...e+poacquirerelease+fencembonceonce.litmus} | 2 +-
11 files changed, 30 insertions(+), 30 deletions(-)
rename tools/memory-model/litmus-tests/{IRIW+mbonceonces+OnceOnce.litmus => IRIW+fencembonceonces+OnceOnce.litmus} (95%)
rename tools/memory-model/litmus-tests/{LB+ctrlonceonce+mbonceonce.litmus => LB+fencembonceonce+ctrlonceonce.litmus} (95%)
rename tools/memory-model/litmus-tests/{MP+wmbonceonce+rmbonceonce.litmus => MP+fencewmbonceonce+fencermbonceonce.litmus} (91%)
rename tools/memory-model/litmus-tests/{R+mbonceonces.litmus => R+fencembonceonces.litmus} (95%)
rename tools/memory-model/litmus-tests/{S+wmbonceonce+poacquireonce.litmus => S+fencewmbonceonce+poacquireonce.litmus} (90%)
rename tools/memory-model/litmus-tests/{SB+mbonceonces.litmus => SB+fencembonceonces.litmus} (95%)
rename tools/memory-model/litmus-tests/{WRC+pooncerelease+rmbonceonce+Once.litmus => WRC+pooncerelease+fencermbonceonce+Once.litmus} (93%)
rename tools/memory-model/litmus-tests/{Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus => Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus} (94%)
diff --git a/tools/memory-model/Documentation/recipes.txt b/tools/memory-model/Documentation/recipes.txt
index 1fea8ef2b184..af72700cc20a 100644
--- a/tools/memory-model/Documentation/recipes.txt
+++ b/tools/memory-model/Documentation/recipes.txt
@@ -126,7 +126,7 @@ However, it is not necessarily the case that accesses ordered by
locking will be seen as ordered by CPUs not holding that lock.
Consider this example:
- /* See Z6.0+pooncelock+pooncelock+pombonce.litmus. */
+ /* See Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus. */
void CPU0(void)
{
spin_lock(&mylock);
@@ -292,7 +292,7 @@ and to use smp_load_acquire() instead of smp_rmb(). However, the older
smp_wmb() and smp_rmb() APIs are still heavily used, so it is important
to understand their use cases. The general approach is shown below:
- /* See MP+wmbonceonce+rmbonceonce.litmus. */
+ /* See MP+fencewmbonceonce+fencermbonceonce.litmus. */
void CPU0(void)
{
WRITE_ONCE(x, 1);
@@ -360,7 +360,7 @@ can be seen in the LB+poonceonces.litmus litmus test.
One way of avoiding the counter-intuitive outcome is through the use of a
control dependency paired with a full memory barrier:
- /* See LB+ctrlonceonce+mbonceonce.litmus. */
+ /* See LB+fencembonceonce+ctrlonceonce.litmus. */
void CPU0(void)
{
r0 = READ_ONCE(x);
@@ -476,7 +476,7 @@ that one CPU first stores to one variable and then loads from a second,
while another CPU stores to the second variable and then loads from the
first. Preserving order requires nothing less than full barriers:
- /* See SB+mbonceonces.litmus. */
+ /* See SB+fencembonceonces.litmus. */
void CPU0(void)
{
WRITE_ONCE(x, 1);
diff --git a/tools/memory-model/README b/tools/memory-model/README
index 734f7feaa5dc..ee987ce20aae 100644
--- a/tools/memory-model/README
+++ b/tools/memory-model/README
@@ -35,13 +35,13 @@ BASIC USAGE: HERD7
The memory model is used, in conjunction with "herd7", to exhaustively
explore the state space of small litmus tests.
-For example, to run SB+mbonceonces.litmus against the memory model:
+For example, to run SB+fencembonceonces.litmus against the memory model:
- $ herd7 -conf linux-kernel.cfg litmus-tests/SB+mbonceonces.litmus
+ $ herd7 -conf linux-kernel.cfg litmus-tests/SB+fencembonceonces.litmus
Here is the corresponding output:
- Test SB+mbonceonces Allowed
+ Test SB+fencembonceonces Allowed
States 3
0:r0=0; 1:r0=1;
0:r0=1; 1:r0=0;
@@ -50,8 +50,8 @@ Here is the corresponding output:
Witnesses
Positive: 0 Negative: 3
Condition exists (0:r0=0 /\ 1:r0=0)
- Observation SB+mbonceonces Never 0 3
- Time SB+mbonceonces 0.01
+ Observation SB+fencembonceonces Never 0 3
+ Time SB+fencembonceonces 0.01
Hash=d66d99523e2cac6b06e66f4c995ebb48
The "Positive: 0 Negative: 3" and the "Never 0 3" each indicate that
@@ -67,16 +67,16 @@ BASIC USAGE: KLITMUS7
The "klitmus7" tool converts a litmus test into a Linux kernel module,
which may then be loaded and run.
-For example, to run SB+mbonceonces.litmus against hardware:
+For example, to run SB+fencembonceonces.litmus against hardware:
$ mkdir mymodules
- $ klitmus7 -o mymodules litmus-tests/SB+mbonceonces.litmus
+ $ klitmus7 -o mymodules litmus-tests/SB+fencembonceonces.litmus
$ cd mymodules ; make
$ sudo sh run.sh
The corresponding output includes:
- Test SB+mbonceonces Allowed
+ Test SB+fencembonceonces Allowed
Histogram (3 states)
644580 :>0:r0=1; 1:r0=0;
644328 :>0:r0=0; 1:r0=1;
@@ -86,8 +86,8 @@ The corresponding output includes:
Positive: 0, Negative: 2000000
Condition exists (0:r0=0 /\ 1:r0=0) is NOT validated
Hash=d66d99523e2cac6b06e66f4c995ebb48
- Observation SB+mbonceonces Never 0 2000000
- Time SB+mbonceonces 0.16
+ Observation SB+fencembonceonces Never 0 2000000
+ Time SB+fencembonceonces 0.16
The "Positive: 0 Negative: 2000000" and the "Never 0 2000000" indicate
that during two million trials, the state specified in this litmus
diff --git a/tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus b/tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus
rename to tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus
index 98a3716efa37..e729d2776e89 100644
--- a/tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus
+++ b/tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus
@@ -1,4 +1,4 @@
-C IRIW+mbonceonces+OnceOnce
+C IRIW+fencembonceonces+OnceOnce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus b/tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus
rename to tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus
index de6708229dd1..4727f5aaf03b 100644
--- a/tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus
+++ b/tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus
@@ -1,4 +1,4 @@
-C LB+ctrlonceonce+mbonceonce
+C LB+fencembonceonce+ctrlonceonce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus b/tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus
similarity index 91%
rename from tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus
rename to tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus
index c078f38ff27a..a273da9faa6d 100644
--- a/tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus
+++ b/tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus
@@ -1,4 +1,4 @@
-C MP+wmbonceonce+rmbonceonce
+C MP+fencewmbonceonce+fencermbonceonce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/R+mbonceonces.litmus b/tools/memory-model/litmus-tests/R+fencembonceonces.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/R+mbonceonces.litmus
rename to tools/memory-model/litmus-tests/R+fencembonceonces.litmus
index a0e884ad2132..222a0b850b4a 100644
--- a/tools/memory-model/litmus-tests/R+mbonceonces.litmus
+++ b/tools/memory-model/litmus-tests/R+fencembonceonces.litmus
@@ -1,4 +1,4 @@
-C R+mbonceonces
+C R+fencembonceonces
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/README b/tools/memory-model/litmus-tests/README
index 00140aaf58b7..4581ec2d3c57 100644
--- a/tools/memory-model/litmus-tests/README
+++ b/tools/memory-model/litmus-tests/README
@@ -18,7 +18,7 @@ CoWW+poonceonce.litmus
Test of write-write coherence, that is, whether or not two
successive writes to the same variable are ordered.
-IRIW+mbonceonces+OnceOnce.litmus
+IRIW+fencembonceonces+OnceOnce.litmus
Test of independent reads from independent writes with smp_mb()
between each pairs of reads. In other words, is smp_mb()
sufficient to cause two different reading processes to agree on
@@ -47,7 +47,7 @@ ISA2+pooncerelease+poacquirerelease+poacquireonce.litmus
Can a release-acquire chain order a prior store against
a later load?
-LB+ctrlonceonce+mbonceonce.litmus
+LB+fencembonceonce+ctrlonceonce.litmus
Does a control dependency and an smp_mb() suffice for the
load-buffering litmus test, where each process reads from one
of two variables then writes to the other?
@@ -88,14 +88,14 @@ MP+porevlocks.litmus
As below, but with the first access of the writer process
and the second access of reader process protected by a lock.
-MP+wmbonceonce+rmbonceonce.litmus
+MP+fencewmbonceonce+fencermbonceonce.litmus
Does a smp_wmb() (between the stores) and an smp_rmb() (between
the loads) suffice for the message-passing litmus test, where one
process writes data and then a flag, and the other process reads
the flag and then the data. (This is similar to the ISA2 tests,
but with two processes instead of three.)
-R+mbonceonces.litmus
+R+fencembonceonces.litmus
This is the fully ordered (via smp_mb()) version of one of
the classic counterintuitive litmus tests that illustrates the
effects of store propagation delays.
@@ -103,7 +103,7 @@ R+mbonceonces.litmus
R+poonceonces.litmus
As above, but without the smp_mb() invocations.
-SB+mbonceonces.litmus
+SB+fencembonceonces.litmus
This is the fully ordered (again, via smp_mb() version of store
buffering, which forms the core of Dekker's mutual-exclusion
algorithm.
@@ -123,12 +123,12 @@ SB+rfionceonce-poonceonces.litmus
S+poonceonces.litmus
As below, but without the smp_wmb() and acquire load.
-S+wmbonceonce+poacquireonce.litmus
+S+fencewmbonceonce+poacquireonce.litmus
Can a smp_wmb(), instead of a release, and an acquire order
a prior store against a subsequent store?
WRC+poonceonces+Once.litmus
-WRC+pooncerelease+rmbonceonce+Once.litmus
+WRC+pooncerelease+fencermbonceonce+Once.litmus
These two are members of an extension of the MP litmus-test
class in which the first write is moved to a separate process.
The second is forbidden because smp_store_release() is
@@ -143,7 +143,7 @@ Z6.0+pooncelock+poonceLock+pombonce.litmus
As above, but with smp_mb__after_spinlock() immediately
following the spin_lock().
-Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus
+Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
Is the ordering provided by a release-acquire chain sufficient
to make ordering apparent to accesses by a process that does
not participate in that release-acquire chain?
diff --git a/tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus b/tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus
similarity index 90%
rename from tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus
rename to tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus
index c53350205d28..18479823cd6c 100644
--- a/tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus
+++ b/tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus
@@ -1,4 +1,4 @@
-C S+wmbonceonce+poacquireonce
+C S+fencewmbonceonce+poacquireonce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/SB+mbonceonces.litmus b/tools/memory-model/litmus-tests/SB+fencembonceonces.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/SB+mbonceonces.litmus
rename to tools/memory-model/litmus-tests/SB+fencembonceonces.litmus
index 74b874ffa8da..ed5fff18d223 100644
--- a/tools/memory-model/litmus-tests/SB+mbonceonces.litmus
+++ b/tools/memory-model/litmus-tests/SB+fencembonceonces.litmus
@@ -1,4 +1,4 @@
-C SB+mbonceonces
+C SB+fencembonceonces
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus b/tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus
similarity index 93%
rename from tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus
rename to tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus
index ad3448b941e6..e9947250d7de 100644
--- a/tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus
+++ b/tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus
@@ -1,4 +1,4 @@
-C WRC+pooncerelease+rmbonceonce+Once
+C WRC+pooncerelease+fencermbonceonce+Once
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus b/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
similarity index 94%
rename from tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus
rename to tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
index a20fc3fafb53..88e70b87a683 100644
--- a/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus
+++ b/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
@@ -1,4 +1,4 @@
-C Z6.0+pooncerelease+poacquirerelease+mbonceonce
+C Z6.0+pooncerelease+poacquirerelease+fencembonceonce
(*
* Result: Sometimes
--
2.17.1
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Add litmus test for full multicopy atomicity
2018-07-16 18:05 ` [PATCH memory-model 01/14] tools/memory-model: Add litmus test for full multicopy atomicity Paul E. McKenney
@ 2018-07-17 9:25 ` tip-bot for Paul E. McKenney
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Paul E. McKenney @ 2018-07-17 9:25 UTC (permalink / raw)
To: linux-tip-commits
Cc: torvalds, stern, mingo, tglx, hpa, paulmck, linux-kernel, peterz,
andrea.parri
Commit-ID: b464818978d45cd4d78c8f13207891142c68bea9
Gitweb: https://git.kernel.org/tip/b464818978d45cd4d78c8f13207891142c68bea9
Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
AuthorDate: Mon, 16 Jul 2018 11:05:52 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:29:29 +0200
tools/memory-model: Add litmus test for full multicopy atomicity
This commit adds a litmus test suggested by Alan Stern that is forbidden
on fully multicopy atomic systems, but allowed on other-multicopy and
on non-multicopy atomic systems. For reference, s390 is fully multicopy
atomic, x86 and ARMv8 are other-multicopy atomic, and ARMv7 and powerpc
are non-multicopy atomic.
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-1-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
tools/memory-model/litmus-tests/README | 9 ++++++
.../litmus-tests/SB+rfionceonce-poonceonces.litmus | 32 ++++++++++++++++++++++
2 files changed, 41 insertions(+)
diff --git a/tools/memory-model/litmus-tests/README b/tools/memory-model/litmus-tests/README
index 17eb9a8c222d..00140aaf58b7 100644
--- a/tools/memory-model/litmus-tests/README
+++ b/tools/memory-model/litmus-tests/README
@@ -111,6 +111,15 @@ SB+mbonceonces.litmus
SB+poonceonces.litmus
As above, but without the smp_mb() invocations.
+SB+rfionceonce-poonceonces.litmus
+ This litmus test demonstrates that LKMM is not fully multicopy
+ atomic. (Neither is it other multicopy atomic.) This litmus test
+ also demonstrates the "locations" debugging aid, which designates
+ additional registers and locations to be printed out in the dump
+ of final states in the herd7 output. Without the "locations"
+ statement, only those registers and locations mentioned in the
+ "exists" clause will be printed.
+
S+poonceonces.litmus
As below, but without the smp_wmb() and acquire load.
diff --git a/tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus b/tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus
new file mode 100644
index 000000000000..04a16603660b
--- /dev/null
+++ b/tools/memory-model/litmus-tests/SB+rfionceonce-poonceonces.litmus
@@ -0,0 +1,32 @@
+C SB+rfionceonce-poonceonces
+
+(*
+ * Result: Sometimes
+ *
+ * This litmus test demonstrates that LKMM is not fully multicopy atomic.
+ *)
+
+{}
+
+P0(int *x, int *y)
+{
+ int r1;
+ int r2;
+
+ WRITE_ONCE(*x, 1);
+ r1 = READ_ONCE(*x);
+ r2 = READ_ONCE(*y);
+}
+
+P1(int *x, int *y)
+{
+ int r3;
+ int r4;
+
+ WRITE_ONCE(*y, 1);
+ r3 = READ_ONCE(*y);
+ r4 = READ_ONCE(*x);
+}
+
+locations [0:r1; 1:r3; x; y] (* Debug aid: Print things not in "exists". *)
+exists (0:r2=0 /\ 1:r4=0)
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name
2018-07-16 18:05 ` [PATCH memory-model 02/14] tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name Paul E. McKenney
@ 2018-07-17 9:26 ` tip-bot for Paul E. McKenney
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Paul E. McKenney @ 2018-07-17 9:26 UTC (permalink / raw)
To: linux-tip-commits
Cc: torvalds, andrea.parri, hpa, stern, mingo, tglx, paulmck, peterz,
linux-kernel
Commit-ID: acb6c96c52ac0da6bb464173ad2cf5ada9049ad4
Gitweb: https://git.kernel.org/tip/acb6c96c52ac0da6bb464173ad2cf5ada9049ad4
Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
AuthorDate: Mon, 16 Jul 2018 11:05:53 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:29:31 +0200
tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name
The names on the first line of the litmus tests are arbitrary,
but the convention is that they be the filename without the trailing
".litmus". This commit therefore removes the stray trailing ".litmus"
from ISA2+pooncelock+pooncelock+pombonce.litmus's name.
Reported-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-2-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
.../litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus b/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus
index 7a39a0aaa976..0f749e419b34 100644
--- a/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus
+++ b/tools/memory-model/litmus-tests/ISA2+pooncelock+pooncelock+pombonce.litmus
@@ -1,4 +1,4 @@
-C ISA2+pooncelock+pooncelock+pombonce.litmus
+C ISA2+pooncelock+pooncelock+pombonce
(*
* Result: Sometimes
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] MAINTAINERS: Add Daniel Lustig as an LKMM reviewer
2018-07-16 18:05 ` [PATCH memory-model 03/14] MAINTAINERS: Add Daniel Lustig as an LKMM reviewer Paul E. McKenney
@ 2018-07-17 9:26 ` tip-bot for Palmer Dabbelt
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Palmer Dabbelt @ 2018-07-17 9:26 UTC (permalink / raw)
To: linux-tip-commits
Cc: hpa, linux-kernel, andrea.parri, torvalds, dlustig, tglx,
paulmck, mingo, peterz, palmer
Commit-ID: 2caf412dc46149b8714e0f52c8337e8eda035a83
Gitweb: https://git.kernel.org/tip/2caf412dc46149b8714e0f52c8337e8eda035a83
Author: Palmer Dabbelt <palmer@sifive.com>
AuthorDate: Mon, 16 Jul 2018 11:05:54 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:29:32 +0200
MAINTAINERS: Add Daniel Lustig as an LKMM reviewer
Dan runs the RISC-V memory model working group. I've been forwarding
him LKMM emails that end up in my inbox, but I'm far from an expert in
this stuff. He requested to be added as a reviewer, which seems sane to
me as it'll take a human out of the loop.
Signed-off-by: Palmer Dabbelt <palmer@sifive.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Daniel Lustig <dlustig@nvidia.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-3-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 192d7f73fd01..eb427596f3e0 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8316,6 +8316,7 @@ M: Jade Alglave <j.alglave@ucl.ac.uk>
M: Luc Maranget <luc.maranget@inria.fr>
M: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
R: Akira Yokosawa <akiyks@gmail.com>
+R: Daniel Lustig <dlustig@nvidia.com>
L: linux-kernel@vger.kernel.org
S: Supported
T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example
2018-07-16 18:05 ` [PATCH memory-model 04/14] locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example Paul E. McKenney
@ 2018-07-17 9:27 ` tip-bot for SeongJae Park
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for SeongJae Park @ 2018-07-17 9:27 UTC (permalink / raw)
To: linux-tip-commits
Cc: sj38.park, linux-kernel, paulmck, mingo, peterz, hpa, torvalds,
byungchul.park, tglx
Commit-ID: 68ec3444b32158541f4bbd47ed9345634c56399d
Gitweb: https://git.kernel.org/tip/68ec3444b32158541f4bbd47ed9345634c56399d
Author: SeongJae Park <sj38.park@gmail.com>
AuthorDate: Mon, 16 Jul 2018 11:05:55 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:29:33 +0200
locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example
Translate this commit to Korean:
5846581e3563 ("locking/memory-barriers.txt: Fix broken DMA vs. MMIO ordering example")
[ paulmck: Updated based on feedback from Byungchul Park. ]
Signed-off-by: SeongJae Park <sj38.park@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Byungchul Park <byungchul.park@lge.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-4-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
.../translations/ko_KR/memory-barriers.txt | 22 +++++++++++-----------
1 file changed, 11 insertions(+), 11 deletions(-)
diff --git a/Documentation/translations/ko_KR/memory-barriers.txt b/Documentation/translations/ko_KR/memory-barriers.txt
index 921739d00f69..7f01fb1c1084 100644
--- a/Documentation/translations/ko_KR/memory-barriers.txt
+++ b/Documentation/translations/ko_KR/memory-barriers.txt
@@ -1891,22 +1891,22 @@ Mandatory 배리어들은 SMP 시스템에서도 UP 시스템에서도 SMP 효
/* 소유권을 수정 */
desc->status = DEVICE_OWN;
- /* MMIO 를 통해 디바이스에 공지를 하기 전에 메모리를 동기화 */
- wmb();
-
/* 업데이트된 디스크립터의 디바이스에 공지 */
writel(DESC_NOTIFY, doorbell);
}
dma_rmb() 는 디스크립터로부터 데이터를 읽어오기 전에 디바이스가 소유권을
- 내놓았음을 보장하게 하고, dma_wmb() 는 디바이스가 자신이 소유권을 다시
- 가졌음을 보기 전에 디스크립터에 데이터가 쓰였음을 보장합니다. wmb() 는
- 캐시 일관성이 없는 (cache incoherent) MMIO 영역에 쓰기를 시도하기 전에
- 캐시 일관성이 있는 메모리 (cache coherent memory) 쓰기가 완료되었음을
- 보장해주기 위해 필요합니다.
-
- consistent memory 에 대한 자세한 내용을 위해선 Documentation/DMA-API.txt
- 문서를 참고하세요.
+ 내려놓았을 것을 보장하고, dma_wmb() 는 디바이스가 자신이 소유권을 다시
+ 가졌음을 보기 전에 디스크립터에 데이터가 쓰였을 것을 보장합니다. 참고로,
+ writel() 을 사용하면 캐시 일관성이 있는 메모리 (cache coherent memory)
+ 쓰기가 MMIO 영역에의 쓰기 전에 완료되었을 것을 보장하므로 writel() 앞에
+ wmb() 를 실행할 필요가 없음을 알아두시기 바랍니다. writel() 보다 비용이
+ 저렴한 writel_relaxed() 는 이런 보장을 제공하지 않으므로 여기선 사용되지
+ 않아야 합니다.
+
+ writel_relaxed() 와 같은 완화된 I/O 접근자들에 대한 자세한 내용을 위해서는
+ "커널 I/O 배리어의 효과" 섹션을, consistent memory 에 대한 자세한 내용을
+ 위해선 Documentation/DMA-API.txt 문서를 참고하세요.
MMIO 쓰기 배리어
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Remove ACCESS_ONCE() from recipes
2018-07-16 18:05 ` [PATCH memory-model 05/14] tools/memory-model: Remove ACCESS_ONCE() from recipes Paul E. McKenney
@ 2018-07-17 9:27 ` tip-bot for Mark Rutland
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Mark Rutland @ 2018-07-17 9:27 UTC (permalink / raw)
To: linux-tip-commits
Cc: luc.maranget, will.deacon, boqun.feng, mark.rutland, tglx,
j.alglave, stern, dhowells, paulmck, hpa, akiyks, torvalds,
linux-kernel, andrea.parri, npiggin, peterz, mingo
Commit-ID: 5bde06b63af2af8c2326270aae041067eef89245
Gitweb: https://git.kernel.org/tip/5bde06b63af2af8c2326270aae041067eef89245
Author: Mark Rutland <mark.rutland@arm.com>
AuthorDate: Mon, 16 Jul 2018 11:05:56 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:29:34 +0200
tools/memory-model: Remove ACCESS_ONCE() from recipes
Since commit:
b899a850431e2dd0 ("compiler.h: Remove ACCESS_ONCE()")
... there has been no definition of ACCESS_ONCE() in the kernel tree,
and it has been necessary to use READ_ONCE() or WRITE_ONCE() instead.
Let's update the exmaples in recipes.txt likewise for consistency, using
READ_ONCE() for reads.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-arch@vger.kernel.org
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180716180605.16115-5-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
tools/memory-model/Documentation/recipes.txt | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/memory-model/Documentation/recipes.txt b/tools/memory-model/Documentation/recipes.txt
index ee4309a87fc4..1fea8ef2b184 100644
--- a/tools/memory-model/Documentation/recipes.txt
+++ b/tools/memory-model/Documentation/recipes.txt
@@ -322,9 +322,9 @@ the following write-side code fragment:
And the xlog_valid_lsn() function in fs/xfs/xfs_log_priv.h contains
the corresponding read-side code fragment:
- cur_cycle = ACCESS_ONCE(log->l_curr_cycle);
+ cur_cycle = READ_ONCE(log->l_curr_cycle);
smp_rmb();
- cur_block = ACCESS_ONCE(log->l_curr_block);
+ cur_block = READ_ONCE(log->l_curr_block);
Alternatively, consider the following comment in function
perf_output_put_handle() in kernel/events/ring_buffer.c:
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Remove ACCESS_ONCE() from model
2018-07-16 18:05 ` [PATCH memory-model 06/14] tools/memory-model: Remove ACCESS_ONCE() from model Paul E. McKenney
@ 2018-07-17 9:28 ` tip-bot for Mark Rutland
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Mark Rutland @ 2018-07-17 9:28 UTC (permalink / raw)
To: linux-tip-commits
Cc: paulmck, dhowells, andrea.parri, stern, mingo, npiggin, akiyks,
peterz, mark.rutland, linux-kernel, hpa, tglx, will.deacon,
luc.maranget, torvalds, j.alglave, boqun.feng
Commit-ID: af41db5ef7c7cf0fee860f65cab59d0d188098d2
Gitweb: https://git.kernel.org/tip/af41db5ef7c7cf0fee860f65cab59d0d188098d2
Author: Mark Rutland <mark.rutland@arm.com>
AuthorDate: Mon, 16 Jul 2018 11:05:57 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:29:35 +0200
tools/memory-model: Remove ACCESS_ONCE() from model
Since commit:
b899a850431e2dd0 ("compiler.h: Remove ACCESS_ONCE()")
... there has been no definition of ACCESS_ONCE() in the kernel tree,
and it has been necessary to use READ_ONCE() or WRITE_ONCE() instead.
Correspondingly, let's remove ACCESS_ONCE() from the kernel memory
model.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-arch@vger.kernel.org
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180716180605.16115-6-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
tools/memory-model/linux-kernel.bell | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/memory-model/linux-kernel.bell b/tools/memory-model/linux-kernel.bell
index 64f5740e0e75..b84fb2f67109 100644
--- a/tools/memory-model/linux-kernel.bell
+++ b/tools/memory-model/linux-kernel.bell
@@ -13,7 +13,7 @@
"Linux-kernel memory consistency model"
-enum Accesses = 'once (*READ_ONCE,WRITE_ONCE,ACCESS_ONCE*) ||
+enum Accesses = 'once (*READ_ONCE,WRITE_ONCE*) ||
'release (*smp_store_release*) ||
'acquire (*smp_load_acquire*) ||
'noreturn (* R of non-return RMW *)
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Make scripts executable
2018-07-16 18:05 ` [PATCH memory-model 07/14] tools/memory-model: Make scripts executable Paul E. McKenney
@ 2018-07-17 9:29 ` tip-bot for Paul E. McKenney
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Paul E. McKenney @ 2018-07-17 9:29 UTC (permalink / raw)
To: linux-tip-commits
Cc: linux-kernel, hpa, akiyks, peterz, tglx, mingo, paulmck, torvalds
Commit-ID: 24675bb554f9e3d0e2bd61f6a7d8da50d224c8ff
Gitweb: https://git.kernel.org/tip/24675bb554f9e3d0e2bd61f6a7d8da50d224c8ff
Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
AuthorDate: Mon, 16 Jul 2018 11:05:58 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:17 +0200
tools/memory-model: Make scripts executable
This commit makes the scripts executable to avoid the need for everyone
to do so manually in their archive.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Akira Yokosawa <akiyks@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-7-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
tools/memory-model/scripts/checkalllitmus.sh | 2 +-
tools/memory-model/scripts/checklitmus.sh | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/memory-model/scripts/checkalllitmus.sh b/tools/memory-model/scripts/checkalllitmus.sh
old mode 100644
new mode 100755
index af0aa15ab84e..ca528f9a24d4
--- a/tools/memory-model/scripts/checkalllitmus.sh
+++ b/tools/memory-model/scripts/checkalllitmus.sh
@@ -9,7 +9,7 @@
# appended.
#
# Usage:
-# sh checkalllitmus.sh [ directory ]
+# checkalllitmus.sh [ directory ]
#
# The LINUX_HERD_OPTIONS environment variable may be used to specify
# arguments to herd, whose default is defined by the checklitmus.sh script.
diff --git a/tools/memory-model/scripts/checklitmus.sh b/tools/memory-model/scripts/checklitmus.sh
old mode 100644
new mode 100755
index e2e477472844..bf12a75c0719
--- a/tools/memory-model/scripts/checklitmus.sh
+++ b/tools/memory-model/scripts/checklitmus.sh
@@ -8,7 +8,7 @@
# with ".out" appended.
#
# Usage:
-# sh checklitmus.sh file.litmus
+# checklitmus.sh file.litmus
#
# The LINUX_HERD_OPTIONS environment variable may be used to specify
# arguments to herd, which default to "-conf linux-kernel.cfg". Thus,
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Add informal LKMM documentation to MAINTAINERS
2018-07-16 18:06 ` [PATCH memory-model 09/14] tools/memory-model: Add informal LKMM documentation to MAINTAINERS Paul E. McKenney
@ 2018-07-17 9:30 ` tip-bot for Paul E. McKenney
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Paul E. McKenney @ 2018-07-17 9:30 UTC (permalink / raw)
To: linux-tip-commits
Cc: tglx, stern, linux-kernel, davem, akiyks, andrea.parri,
luc.maranget, boqun.feng, npiggin, hpa, paulmck, dhowells,
dlustig, j.alglave, mingo, will.deacon, torvalds, peterz
Commit-ID: 70b83069f70d185356beba3202b9d167ee39f051
Gitweb: https://git.kernel.org/tip/70b83069f70d185356beba3202b9d167ee39f051
Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
AuthorDate: Mon, 16 Jul 2018 11:06:00 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:32 +0200
tools/memory-model: Add informal LKMM documentation to MAINTAINERS
The Linux-kernel memory model has been informal, with a number of
text files documenting it. It would be good to make sure that these
informal descriptions are kept up to date and/or pruned appropriately.
This commit therefore brings more of those text files into the LKMM
MAINTAINERS file entry.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Daniel Lustig <dlustig@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-arch@vger.kernel.org
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180716180605.16115-9-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
MAINTAINERS | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index eb427596f3e0..900889a47627 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8318,9 +8318,14 @@ M: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
R: Akira Yokosawa <akiyks@gmail.com>
R: Daniel Lustig <dlustig@nvidia.com>
L: linux-kernel@vger.kernel.org
+L: linux-arch@vger.kernel.org
S: Supported
T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
F: tools/memory-model/
+F: Documentation/atomic_bitops.txt
+F: Documentation/atomic_t.txt
+F: Documentation/core-api/atomic_ops.rst
+F: Documentation/core-api/refcount-vs-atomic.rst
F: Documentation/memory-barriers.txt
LINUX SECURITY MODULE (LSM) FRAMEWORK
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] sched/core: Use smp_mb() in wake_woken_function()
2018-07-16 18:06 ` [PATCH memory-model 10/14] sched: Use smp_mb() in wake_woken_function() Paul E. McKenney
@ 2018-07-17 9:30 ` tip-bot for Andrea Parri
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Andrea Parri @ 2018-07-17 9:30 UTC (permalink / raw)
To: linux-tip-commits
Cc: linux-kernel, peterz, paulmck, torvalds, andrea.parri, hpa, mingo, tglx
Commit-ID: 76e079fefc8f62bd9b2cd2950814d1ee806e31a5
Gitweb: https://git.kernel.org/tip/76e079fefc8f62bd9b2cd2950814d1ee806e31a5
Author: Andrea Parri <andrea.parri@amarulasolutions.com>
AuthorDate: Mon, 16 Jul 2018 11:06:01 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:33 +0200
sched/core: Use smp_mb() in wake_woken_function()
wake_woken_function() synchronizes with wait_woken() as follows:
[wait_woken] [wake_woken_function]
entry->flags &= ~wq_flag_woken; condition = true;
smp_mb(); smp_wmb();
if (condition) wq_entry->flags |= wq_flag_woken;
break;
This commit replaces the above smp_wmb() with an smp_mb() in order to
guarantee that either wait_woken() sees the wait condition being true
or the store to wq_entry->flags in woken_wake_function() follows the
store in wait_woken() in the coherence order (so that the former can
eventually be observed by wait_woken()).
The commit also fixes a comment associated to set_current_state() in
wait_woken(): the comment pairs the barrier in set_current_state() to
the above smp_wmb(), while the actual pairing involves the barrier in
set_current_state() and the barrier executed by the try_to_wake_up()
in wake_woken_function().
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-10-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
kernel/sched/wait.c | 47 +++++++++++++++++++++--------------------------
1 file changed, 21 insertions(+), 26 deletions(-)
diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c
index 928be527477e..a7a2aaa3026a 100644
--- a/kernel/sched/wait.c
+++ b/kernel/sched/wait.c
@@ -392,35 +392,36 @@ static inline bool is_kthread_should_stop(void)
* if (condition)
* break;
*
- * p->state = mode; condition = true;
- * smp_mb(); // A smp_wmb(); // C
- * if (!wq_entry->flags & WQ_FLAG_WOKEN) wq_entry->flags |= WQ_FLAG_WOKEN;
- * schedule() try_to_wake_up();
- * p->state = TASK_RUNNING; ~~~~~~~~~~~~~~~~~~
- * wq_entry->flags &= ~WQ_FLAG_WOKEN; condition = true;
- * smp_mb() // B smp_wmb(); // C
- * wq_entry->flags |= WQ_FLAG_WOKEN;
- * }
- * remove_wait_queue(&wq_head, &wait);
+ * // in wait_woken() // in woken_wake_function()
*
+ * p->state = mode; wq_entry->flags |= WQ_FLAG_WOKEN;
+ * smp_mb(); // A try_to_wake_up():
+ * if (!(wq_entry->flags & WQ_FLAG_WOKEN)) <full barrier>
+ * schedule() if (p->state & mode)
+ * p->state = TASK_RUNNING; p->state = TASK_RUNNING;
+ * wq_entry->flags &= ~WQ_FLAG_WOKEN; ~~~~~~~~~~~~~~~~~~
+ * smp_mb(); // B condition = true;
+ * } smp_mb(); // C
+ * remove_wait_queue(&wq_head, &wait); wq_entry->flags |= WQ_FLAG_WOKEN;
*/
long wait_woken(struct wait_queue_entry *wq_entry, unsigned mode, long timeout)
{
- set_current_state(mode); /* A */
/*
- * The above implies an smp_mb(), which matches with the smp_wmb() from
- * woken_wake_function() such that if we observe WQ_FLAG_WOKEN we must
- * also observe all state before the wakeup.
+ * The below executes an smp_mb(), which matches with the full barrier
+ * executed by the try_to_wake_up() in woken_wake_function() such that
+ * either we see the store to wq_entry->flags in woken_wake_function()
+ * or woken_wake_function() sees our store to current->state.
*/
+ set_current_state(mode); /* A */
if (!(wq_entry->flags & WQ_FLAG_WOKEN) && !is_kthread_should_stop())
timeout = schedule_timeout(timeout);
__set_current_state(TASK_RUNNING);
/*
- * The below implies an smp_mb(), it too pairs with the smp_wmb() from
- * woken_wake_function() such that we must either observe the wait
- * condition being true _OR_ WQ_FLAG_WOKEN such that we will not miss
- * an event.
+ * The below executes an smp_mb(), which matches with the smp_mb() (C)
+ * in woken_wake_function() such that either we see the wait condition
+ * being true or the store to wq_entry->flags in woken_wake_function()
+ * follows ours in the coherence order.
*/
smp_store_mb(wq_entry->flags, wq_entry->flags & ~WQ_FLAG_WOKEN); /* B */
@@ -430,14 +431,8 @@ EXPORT_SYMBOL(wait_woken);
int woken_wake_function(struct wait_queue_entry *wq_entry, unsigned mode, int sync, void *key)
{
- /*
- * Although this function is called under waitqueue lock, LOCK
- * doesn't imply write barrier and the users expects write
- * barrier semantics on wakeup functions. The following
- * smp_wmb() is equivalent to smp_wmb() in try_to_wake_up()
- * and is paired with smp_store_mb() in wait_woken().
- */
- smp_wmb(); /* C */
+ /* Pairs with the smp_store_mb() in wait_woken(). */
+ smp_mb(); /* C */
wq_entry->flags |= WQ_FLAG_WOKEN;
return default_wake_function(wq_entry, mode, sync, key);
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] locking/spinlock, sched/core: Clarify requirements for smp_mb__after_spinlock()
2018-07-16 18:06 ` [PATCH memory-model 11/14] locking: Clarify requirements for smp_mb__after_spinlock() Paul E. McKenney
@ 2018-07-17 9:31 ` tip-bot for Andrea Parri
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Andrea Parri @ 2018-07-17 9:31 UTC (permalink / raw)
To: linux-tip-commits
Cc: linux-kernel, andrea.parri, torvalds, mingo, boqun.feng, paulmck,
tglx, will.deacon, peterz, hpa
Commit-ID: 3d85b2703783636366560c94842affd8608ec9d1
Gitweb: https://git.kernel.org/tip/3d85b2703783636366560c94842affd8608ec9d1
Author: Andrea Parri <andrea.parri@amarulasolutions.com>
AuthorDate: Mon, 16 Jul 2018 11:06:02 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:33 +0200
locking/spinlock, sched/core: Clarify requirements for smp_mb__after_spinlock()
There are 11 interpretations of the requirements described in the header
comment for smp_mb__after_spinlock(): one for each LKMM maintainer, and
one currently encoded in the Cat file. Stick to the latter (until a more
satisfactory solution is available).
This also reworks some snippets related to the barrier to illustrate the
requirements and to link them to the idioms which are relied upon at its
call sites.
Suggested-by: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: akiyks@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: stern@rowland.harvard.edu
Link: http://lkml.kernel.org/r/20180716180605.16115-11-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
include/linux/spinlock.h | 53 ++++++++++++++++++++++++++++++++----------------
kernel/sched/core.c | 41 +++++++++++++++++++------------------
2 files changed, 57 insertions(+), 37 deletions(-)
diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
index fd57888d4942..3190997df9ca 100644
--- a/include/linux/spinlock.h
+++ b/include/linux/spinlock.h
@@ -114,29 +114,48 @@ do { \
#endif /*arch_spin_is_contended*/
/*
- * This barrier must provide two things:
+ * smp_mb__after_spinlock() provides the equivalent of a full memory barrier
+ * between program-order earlier lock acquisitions and program-order later
+ * memory accesses.
*
- * - it must guarantee a STORE before the spin_lock() is ordered against a
- * LOAD after it, see the comments at its two usage sites.
+ * This guarantees that the following two properties hold:
*
- * - it must ensure the critical section is RCsc.
+ * 1) Given the snippet:
*
- * The latter is important for cases where we observe values written by other
- * CPUs in spin-loops, without barriers, while being subject to scheduling.
+ * { X = 0; Y = 0; }
*
- * CPU0 CPU1 CPU2
+ * CPU0 CPU1
*
- * for (;;) {
- * if (READ_ONCE(X))
- * break;
- * }
- * X=1
- * <sched-out>
- * <sched-in>
- * r = X;
+ * WRITE_ONCE(X, 1); WRITE_ONCE(Y, 1);
+ * spin_lock(S); smp_mb();
+ * smp_mb__after_spinlock(); r1 = READ_ONCE(X);
+ * r0 = READ_ONCE(Y);
+ * spin_unlock(S);
*
- * without transitivity it could be that CPU1 observes X!=0 breaks the loop,
- * we get migrated and CPU2 sees X==0.
+ * it is forbidden that CPU0 does not observe CPU1's store to Y (r0 = 0)
+ * and CPU1 does not observe CPU0's store to X (r1 = 0); see the comments
+ * preceding the call to smp_mb__after_spinlock() in __schedule() and in
+ * try_to_wake_up().
+ *
+ * 2) Given the snippet:
+ *
+ * { X = 0; Y = 0; }
+ *
+ * CPU0 CPU1 CPU2
+ *
+ * spin_lock(S); spin_lock(S); r1 = READ_ONCE(Y);
+ * WRITE_ONCE(X, 1); smp_mb__after_spinlock(); smp_rmb();
+ * spin_unlock(S); r0 = READ_ONCE(X); r2 = READ_ONCE(X);
+ * WRITE_ONCE(Y, 1);
+ * spin_unlock(S);
+ *
+ * it is forbidden that CPU0's critical section executes before CPU1's
+ * critical section (r0 = 1), CPU2 observes CPU1's store to Y (r1 = 1)
+ * and CPU2 does not observe CPU0's store to X (r2 = 0); see the comments
+ * preceding the calls to smp_rmb() in try_to_wake_up() for similar
+ * snippets but "projected" onto two CPUs.
+ *
+ * Property (2) upgrades the lock to an RCsc lock.
*
* Since most load-store architectures implement ACQUIRE with an smp_mb() after
* the LL/SC loop, they need no further barriers. Similarly all our TSO
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index fe365c9a08e9..0c5ec2abdf93 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1998,21 +1998,20 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
* be possible to, falsely, observe p->on_rq == 0 and get stuck
* in smp_cond_load_acquire() below.
*
- * sched_ttwu_pending() try_to_wake_up()
- * [S] p->on_rq = 1; [L] P->state
- * UNLOCK rq->lock -----.
- * \
- * +--- RMB
- * schedule() /
- * LOCK rq->lock -----'
- * UNLOCK rq->lock
+ * sched_ttwu_pending() try_to_wake_up()
+ * STORE p->on_rq = 1 LOAD p->state
+ * UNLOCK rq->lock
+ *
+ * __schedule() (switch to task 'p')
+ * LOCK rq->lock smp_rmb();
+ * smp_mb__after_spinlock();
+ * UNLOCK rq->lock
*
* [task p]
- * [S] p->state = UNINTERRUPTIBLE [L] p->on_rq
+ * STORE p->state = UNINTERRUPTIBLE LOAD p->on_rq
*
- * Pairs with the UNLOCK+LOCK on rq->lock from the
- * last wakeup of our task and the schedule that got our task
- * current.
+ * Pairs with the LOCK+smp_mb__after_spinlock() on rq->lock in
+ * __schedule(). See the comment for smp_mb__after_spinlock().
*/
smp_rmb();
if (p->on_rq && ttwu_remote(p, wake_flags))
@@ -2026,15 +2025,17 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
* One must be running (->on_cpu == 1) in order to remove oneself
* from the runqueue.
*
- * [S] ->on_cpu = 1; [L] ->on_rq
- * UNLOCK rq->lock
- * RMB
- * LOCK rq->lock
- * [S] ->on_rq = 0; [L] ->on_cpu
+ * __schedule() (switch to task 'p') try_to_wake_up()
+ * STORE p->on_cpu = 1 LOAD p->on_rq
+ * UNLOCK rq->lock
+ *
+ * __schedule() (put 'p' to sleep)
+ * LOCK rq->lock smp_rmb();
+ * smp_mb__after_spinlock();
+ * STORE p->on_rq = 0 LOAD p->on_cpu
*
- * Pairs with the full barrier implied in the UNLOCK+LOCK on rq->lock
- * from the consecutive calls to schedule(); the first switching to our
- * task, the second putting it to sleep.
+ * Pairs with the LOCK+smp_mb__after_spinlock() on rq->lock in
+ * __schedule(). See the comment for smp_mb__after_spinlock().
*/
smp_rmb();
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] sched/Documentation: Update wake_up() & co. memory-barrier guarantees
2018-07-16 18:06 ` [PATCH memory-model 12/14] doc: Update wake_up() & co. memory-barrier guarantees Paul E. McKenney
@ 2018-07-17 9:31 ` tip-bot for Andrea Parri
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Andrea Parri @ 2018-07-17 9:31 UTC (permalink / raw)
To: linux-tip-commits
Cc: dhowells, torvalds, tglx, stern, paulmck, boqun.feng, peterz,
linux-kernel, j.alglave, dlustig, hpa, mingo, npiggin,
will.deacon, andrea.parri, akiyks, luc.maranget, corbet
Commit-ID: 7696f9910a9a40b8a952f57d3428515fabd2d889
Gitweb: https://git.kernel.org/tip/7696f9910a9a40b8a952f57d3428515fabd2d889
Author: Andrea Parri <andrea.parri@amarulasolutions.com>
AuthorDate: Mon, 16 Jul 2018 11:06:03 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:34 +0200
sched/Documentation: Update wake_up() & co. memory-barrier guarantees
Both the implementation and the users' expectation [1] for the various
wakeup primitives have evolved over time, but the documentation has not
kept up with these changes: brings it into 2018.
[1] http://lkml.kernel.org/r/20180424091510.GB4064@hirez.programming.kicks-ass.net
Also applied feedback from Alan Stern.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Daniel Lustig <dlustig@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-arch@vger.kernel.org
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180716180605.16115-12-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
Documentation/memory-barriers.txt | 43 ++++++++++++++++++++++++---------------
include/linux/sched.h | 4 ++--
kernel/sched/completion.c | 8 ++++----
kernel/sched/core.c | 30 +++++++++++----------------
kernel/sched/wait.c | 8 ++++----
5 files changed, 49 insertions(+), 44 deletions(-)
diff --git a/Documentation/memory-barriers.txt b/Documentation/memory-barriers.txt
index a02d6bbfc9d0..0d8d7ef131e9 100644
--- a/Documentation/memory-barriers.txt
+++ b/Documentation/memory-barriers.txt
@@ -2179,32 +2179,41 @@ or:
event_indicated = 1;
wake_up_process(event_daemon);
-A write memory barrier is implied by wake_up() and co. if and only if they
-wake something up. The barrier occurs before the task state is cleared, and so
-sits between the STORE to indicate the event and the STORE to set TASK_RUNNING:
+A general memory barrier is executed by wake_up() if it wakes something up.
+If it doesn't wake anything up then a memory barrier may or may not be
+executed; you must not rely on it. The barrier occurs before the task state
+is accessed, in particular, it sits between the STORE to indicate the event
+and the STORE to set TASK_RUNNING:
- CPU 1 CPU 2
+ CPU 1 (Sleeper) CPU 2 (Waker)
=============================== ===============================
set_current_state(); STORE event_indicated
smp_store_mb(); wake_up();
- STORE current->state <write barrier>
- <general barrier> STORE current->state
- LOAD event_indicated
+ STORE current->state ...
+ <general barrier> <general barrier>
+ LOAD event_indicated if ((LOAD task->state) & TASK_NORMAL)
+ STORE task->state
-To repeat, this write memory barrier is present if and only if something
-is actually awakened. To see this, consider the following sequence of
-events, where X and Y are both initially zero:
+where "task" is the thread being woken up and it equals CPU 1's "current".
+
+To repeat, a general memory barrier is guaranteed to be executed by wake_up()
+if something is actually awakened, but otherwise there is no such guarantee.
+To see this, consider the following sequence of events, where X and Y are both
+initially zero:
CPU 1 CPU 2
=============================== ===============================
- X = 1; STORE event_indicated
+ X = 1; Y = 1;
smp_mb(); wake_up();
- Y = 1; wait_event(wq, Y == 1);
- wake_up(); load from Y sees 1, no memory barrier
- load from X might see 0
+ LOAD Y LOAD X
+
+If a wakeup does occur, one (at least) of the two loads must see 1. If, on
+the other hand, a wakeup does not occur, both loads might see 0.
-In contrast, if a wakeup does occur, CPU 2's load from X would be guaranteed
-to see 1.
+wake_up_process() always executes a general memory barrier. The barrier again
+occurs before the task state is accessed. In particular, if the wake_up() in
+the previous snippet were replaced by a call to wake_up_process() then one of
+the two loads would be guaranteed to see 1.
The available waker functions include:
@@ -2224,6 +2233,8 @@ The available waker functions include:
wake_up_poll();
wake_up_process();
+In terms of memory ordering, these functions all provide the same guarantees of
+a wake_up() (or stronger).
[!] Note that the memory barriers implied by the sleeper and the waker do _not_
order multiple stores before the wake-up with respect to loads of those stored
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 43731fe51c97..05cd419f962d 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -167,8 +167,8 @@ struct task_group;
* need_sleep = false;
* wake_up_state(p, TASK_UNINTERRUPTIBLE);
*
- * Where wake_up_state() (and all other wakeup primitives) imply enough
- * barriers to order the store of the variable against wakeup.
+ * where wake_up_state() executes a full memory barrier before accessing the
+ * task state.
*
* Wakeup will do: if (@state & p->state) p->state = TASK_RUNNING, that is,
* once it observes the TASK_UNINTERRUPTIBLE store the waking CPU can issue a
diff --git a/kernel/sched/completion.c b/kernel/sched/completion.c
index e426b0cb9ac6..a1ad5b7d5521 100644
--- a/kernel/sched/completion.c
+++ b/kernel/sched/completion.c
@@ -22,8 +22,8 @@
*
* See also complete_all(), wait_for_completion() and related routines.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*/
void complete(struct completion *x)
{
@@ -44,8 +44,8 @@ EXPORT_SYMBOL(complete);
*
* This will wake up all threads waiting on this particular completion event.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*
* Since complete_all() sets the completion of @x permanently to done
* to allow multiple waiters to finish, a call to reinit_completion()
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 0c5ec2abdf93..a0065c84e73f 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -412,8 +412,8 @@ void wake_q_add(struct wake_q_head *head, struct task_struct *task)
* its already queued (either by us or someone else) and will get the
* wakeup due to that.
*
- * This cmpxchg() implies a full barrier, which pairs with the write
- * barrier implied by the wakeup in wake_up_q().
+ * This cmpxchg() executes a full barrier, which pairs with the full
+ * barrier executed by the wakeup in wake_up_q().
*/
if (cmpxchg(&node->next, NULL, WAKE_Q_TAIL))
return;
@@ -441,8 +441,8 @@ void wake_up_q(struct wake_q_head *head)
task->wake_q.next = NULL;
/*
- * wake_up_process() implies a wmb() to pair with the queueing
- * in wake_q_add() so as not to miss wakeups.
+ * wake_up_process() executes a full barrier, which pairs with
+ * the queueing in wake_q_add() so as not to miss wakeups.
*/
wake_up_process(task);
put_task_struct(task);
@@ -1879,8 +1879,7 @@ static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags)
* rq(c1)->lock (if not at the same time, then in that order).
* C) LOCK of the rq(c1)->lock scheduling in task
*
- * Transitivity guarantees that B happens after A and C after B.
- * Note: we only require RCpc transitivity.
+ * Release/acquire chaining guarantees that B happens after A and C after B.
* Note: the CPU doing B need not be c0 or c1
*
* Example:
@@ -1942,16 +1941,9 @@ static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags)
* UNLOCK rq(0)->lock
*
*
- * However; for wakeups there is a second guarantee we must provide, namely we
- * must observe the state that lead to our wakeup. That is, not only must our
- * task observe its own prior state, it must also observe the stores prior to
- * its wakeup.
- *
- * This means that any means of doing remote wakeups must order the CPU doing
- * the wakeup against the CPU the task is going to end up running on. This,
- * however, is already required for the regular Program-Order guarantee above,
- * since the waking CPU is the one issueing the ACQUIRE (smp_cond_load_acquire).
- *
+ * However, for wakeups there is a second guarantee we must provide, namely we
+ * must ensure that CONDITION=1 done by the caller can not be reordered with
+ * accesses to the task state; see try_to_wake_up() and set_current_state().
*/
/**
@@ -1967,6 +1959,9 @@ static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags)
* Atomic against schedule() which would dequeue a task, also see
* set_current_state().
*
+ * This function executes a full memory barrier before accessing the task
+ * state; see set_current_state().
+ *
* Return: %true if @p->state changes (an actual wakeup was done),
* %false otherwise.
*/
@@ -2141,8 +2136,7 @@ out:
*
* Return: 1 if the process was woken up, 0 if it was already running.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * This function executes a full memory barrier before accessing the task state.
*/
int wake_up_process(struct task_struct *p)
{
diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c
index a7a2aaa3026a..870f97b313e3 100644
--- a/kernel/sched/wait.c
+++ b/kernel/sched/wait.c
@@ -134,8 +134,8 @@ static void __wake_up_common_lock(struct wait_queue_head *wq_head, unsigned int
* @nr_exclusive: how many wake-one or wake-many threads to wake up
* @key: is directly passed to the wakeup function
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*/
void __wake_up(struct wait_queue_head *wq_head, unsigned int mode,
int nr_exclusive, void *key)
@@ -180,8 +180,8 @@ EXPORT_SYMBOL_GPL(__wake_up_locked_key_bookmark);
*
* On UP it can prevent extra preemption.
*
- * It may be assumed that this function implies a write memory barrier before
- * changing the task state if and only if any tasks are woken up.
+ * If this function wakes up a task, it executes a full memory barrier before
+ * accessing the task state.
*/
void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode,
int nr_exclusive, void *key)
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model/Documentation: Fix typo, smb->smp
2018-07-16 18:06 ` [PATCH memory-model 13/14] memory-model/Documentation: Fix typo, smb->smp Paul E. McKenney
@ 2018-07-17 9:32 ` tip-bot for Yauheni Kaliuta
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Yauheni Kaliuta @ 2018-07-17 9:32 UTC (permalink / raw)
To: linux-tip-commits
Cc: mingo, peterz, stern, yauheni.kaliuta, hpa, paulmck, tglx,
linux-kernel, torvalds
Commit-ID: 0fcff1715bec7593a0ba86f3fef46cd89af37a8b
Gitweb: https://git.kernel.org/tip/0fcff1715bec7593a0ba86f3fef46cd89af37a8b
Author: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
AuthorDate: Mon, 16 Jul 2018 11:06:04 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:35 +0200
tools/memory-model/Documentation: Fix typo, smb->smp
The tools/memory-model/Documentation/explanation.txt file says
"For each other CPU C', smb_wmb() forces all po-earlier stores"
This commit therefore replaces the "smb_wmb()" with "smp_wmb()".
Signed-off-by: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akiyks@gmail.com
Cc: boqun.feng@gmail.com
Cc: dhowells@redhat.com
Cc: j.alglave@ucl.ac.uk
Cc: linux-arch@vger.kernel.org
Cc: luc.maranget@inria.fr
Cc: npiggin@gmail.com
Cc: parri.andrea@gmail.com
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/20180716180605.16115-13-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
tools/memory-model/Documentation/explanation.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/memory-model/Documentation/explanation.txt b/tools/memory-model/Documentation/explanation.txt
index 1b09f3175a1f..0cbd1ef8f86d 100644
--- a/tools/memory-model/Documentation/explanation.txt
+++ b/tools/memory-model/Documentation/explanation.txt
@@ -804,7 +804,7 @@ type of fence:
Second, some types of fence affect the way the memory subsystem
propagates stores. When a fence instruction is executed on CPU C:
- For each other CPU C', smb_wmb() forces all po-earlier stores
+ For each other CPU C', smp_wmb() forces all po-earlier stores
on C to propagate to C' before any po-later stores do.
For each other CPU C', any store which propagates to C before
^ permalink raw reply related [flat|nested] 28+ messages in thread
* [tip:locking/core] tools/memory-model: Rename litmus tests to comply to norm7
2018-07-16 18:06 ` [PATCH memory-model 14/14] tools/memory-model: Rename litmus tests to comply to norm7 Paul E. McKenney
@ 2018-07-17 9:33 ` tip-bot for Andrea Parri
0 siblings, 0 replies; 28+ messages in thread
From: tip-bot for Andrea Parri @ 2018-07-17 9:33 UTC (permalink / raw)
To: linux-tip-commits
Cc: linux-kernel, stern, npiggin, luc.maranget, paulmck, mingo, hpa,
andrea.parri, tglx, j.alglave, dhowells, torvalds, peterz,
boqun.feng, akiyks, will.deacon
Commit-ID: 71b7ff5ebc9b1d5aa95eb48d6388234f1304fd19
Gitweb: https://git.kernel.org/tip/71b7ff5ebc9b1d5aa95eb48d6388234f1304fd19
Author: Andrea Parri <andrea.parri@amarulasolutions.com>
AuthorDate: Mon, 16 Jul 2018 11:06:05 -0700
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 17 Jul 2018 09:30:36 +0200
tools/memory-model: Rename litmus tests to comply to norm7
norm7 produces the 'normalized' name of a litmus test, when the test
can be generated from a single cycle that passes through each process
exactly once. The commit renames such tests in order to comply to the
naming scheme implemented by this tool.
Signed-off-by: Andrea Parri <andrea.parri@amarulasolutions.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Akira Yokosawa <akiyks@gmail.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Jade Alglave <j.alglave@ucl.ac.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luc Maranget <luc.maranget@inria.fr>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will.deacon@arm.com>
Cc: linux-arch@vger.kernel.org
Cc: parri.andrea@gmail.com
Link: http://lkml.kernel.org/r/20180716180605.16115-14-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
tools/memory-model/Documentation/recipes.txt | 8 ++++----
tools/memory-model/README | 20 ++++++++++----------
....litmus => IRIW+fencembonceonces+OnceOnce.litmus} | 2 +-
...litmus => LB+fencembonceonce+ctrlonceonce.litmus} | 2 +-
...s => MP+fencewmbonceonce+fencermbonceonce.litmus} | 2 +-
...+mbonceonces.litmus => R+fencembonceonces.litmus} | 2 +-
tools/memory-model/litmus-tests/README | 16 ++++++++--------
...itmus => S+fencewmbonceonce+poacquireonce.litmus} | 2 +-
...mbonceonces.litmus => SB+fencembonceonces.litmus} | 2 +-
...> WRC+pooncerelease+fencermbonceonce+Once.litmus} | 2 +-
...erelease+poacquirerelease+fencembonceonce.litmus} | 2 +-
11 files changed, 30 insertions(+), 30 deletions(-)
diff --git a/tools/memory-model/Documentation/recipes.txt b/tools/memory-model/Documentation/recipes.txt
index 1fea8ef2b184..af72700cc20a 100644
--- a/tools/memory-model/Documentation/recipes.txt
+++ b/tools/memory-model/Documentation/recipes.txt
@@ -126,7 +126,7 @@ However, it is not necessarily the case that accesses ordered by
locking will be seen as ordered by CPUs not holding that lock.
Consider this example:
- /* See Z6.0+pooncelock+pooncelock+pombonce.litmus. */
+ /* See Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus. */
void CPU0(void)
{
spin_lock(&mylock);
@@ -292,7 +292,7 @@ and to use smp_load_acquire() instead of smp_rmb(). However, the older
smp_wmb() and smp_rmb() APIs are still heavily used, so it is important
to understand their use cases. The general approach is shown below:
- /* See MP+wmbonceonce+rmbonceonce.litmus. */
+ /* See MP+fencewmbonceonce+fencermbonceonce.litmus. */
void CPU0(void)
{
WRITE_ONCE(x, 1);
@@ -360,7 +360,7 @@ can be seen in the LB+poonceonces.litmus litmus test.
One way of avoiding the counter-intuitive outcome is through the use of a
control dependency paired with a full memory barrier:
- /* See LB+ctrlonceonce+mbonceonce.litmus. */
+ /* See LB+fencembonceonce+ctrlonceonce.litmus. */
void CPU0(void)
{
r0 = READ_ONCE(x);
@@ -476,7 +476,7 @@ that one CPU first stores to one variable and then loads from a second,
while another CPU stores to the second variable and then loads from the
first. Preserving order requires nothing less than full barriers:
- /* See SB+mbonceonces.litmus. */
+ /* See SB+fencembonceonces.litmus. */
void CPU0(void)
{
WRITE_ONCE(x, 1);
diff --git a/tools/memory-model/README b/tools/memory-model/README
index 734f7feaa5dc..ee987ce20aae 100644
--- a/tools/memory-model/README
+++ b/tools/memory-model/README
@@ -35,13 +35,13 @@ BASIC USAGE: HERD7
The memory model is used, in conjunction with "herd7", to exhaustively
explore the state space of small litmus tests.
-For example, to run SB+mbonceonces.litmus against the memory model:
+For example, to run SB+fencembonceonces.litmus against the memory model:
- $ herd7 -conf linux-kernel.cfg litmus-tests/SB+mbonceonces.litmus
+ $ herd7 -conf linux-kernel.cfg litmus-tests/SB+fencembonceonces.litmus
Here is the corresponding output:
- Test SB+mbonceonces Allowed
+ Test SB+fencembonceonces Allowed
States 3
0:r0=0; 1:r0=1;
0:r0=1; 1:r0=0;
@@ -50,8 +50,8 @@ Here is the corresponding output:
Witnesses
Positive: 0 Negative: 3
Condition exists (0:r0=0 /\ 1:r0=0)
- Observation SB+mbonceonces Never 0 3
- Time SB+mbonceonces 0.01
+ Observation SB+fencembonceonces Never 0 3
+ Time SB+fencembonceonces 0.01
Hash=d66d99523e2cac6b06e66f4c995ebb48
The "Positive: 0 Negative: 3" and the "Never 0 3" each indicate that
@@ -67,16 +67,16 @@ BASIC USAGE: KLITMUS7
The "klitmus7" tool converts a litmus test into a Linux kernel module,
which may then be loaded and run.
-For example, to run SB+mbonceonces.litmus against hardware:
+For example, to run SB+fencembonceonces.litmus against hardware:
$ mkdir mymodules
- $ klitmus7 -o mymodules litmus-tests/SB+mbonceonces.litmus
+ $ klitmus7 -o mymodules litmus-tests/SB+fencembonceonces.litmus
$ cd mymodules ; make
$ sudo sh run.sh
The corresponding output includes:
- Test SB+mbonceonces Allowed
+ Test SB+fencembonceonces Allowed
Histogram (3 states)
644580 :>0:r0=1; 1:r0=0;
644328 :>0:r0=0; 1:r0=1;
@@ -86,8 +86,8 @@ The corresponding output includes:
Positive: 0, Negative: 2000000
Condition exists (0:r0=0 /\ 1:r0=0) is NOT validated
Hash=d66d99523e2cac6b06e66f4c995ebb48
- Observation SB+mbonceonces Never 0 2000000
- Time SB+mbonceonces 0.16
+ Observation SB+fencembonceonces Never 0 2000000
+ Time SB+fencembonceonces 0.16
The "Positive: 0 Negative: 2000000" and the "Never 0 2000000" indicate
that during two million trials, the state specified in this litmus
diff --git a/tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus b/tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus
rename to tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus
index 98a3716efa37..e729d2776e89 100644
--- a/tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus
+++ b/tools/memory-model/litmus-tests/IRIW+fencembonceonces+OnceOnce.litmus
@@ -1,4 +1,4 @@
-C IRIW+mbonceonces+OnceOnce
+C IRIW+fencembonceonces+OnceOnce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus b/tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus
rename to tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus
index de6708229dd1..4727f5aaf03b 100644
--- a/tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus
+++ b/tools/memory-model/litmus-tests/LB+fencembonceonce+ctrlonceonce.litmus
@@ -1,4 +1,4 @@
-C LB+ctrlonceonce+mbonceonce
+C LB+fencembonceonce+ctrlonceonce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus b/tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus
similarity index 91%
rename from tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus
rename to tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus
index c078f38ff27a..a273da9faa6d 100644
--- a/tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus
+++ b/tools/memory-model/litmus-tests/MP+fencewmbonceonce+fencermbonceonce.litmus
@@ -1,4 +1,4 @@
-C MP+wmbonceonce+rmbonceonce
+C MP+fencewmbonceonce+fencermbonceonce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/R+mbonceonces.litmus b/tools/memory-model/litmus-tests/R+fencembonceonces.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/R+mbonceonces.litmus
rename to tools/memory-model/litmus-tests/R+fencembonceonces.litmus
index a0e884ad2132..222a0b850b4a 100644
--- a/tools/memory-model/litmus-tests/R+mbonceonces.litmus
+++ b/tools/memory-model/litmus-tests/R+fencembonceonces.litmus
@@ -1,4 +1,4 @@
-C R+mbonceonces
+C R+fencembonceonces
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/README b/tools/memory-model/litmus-tests/README
index 00140aaf58b7..4581ec2d3c57 100644
--- a/tools/memory-model/litmus-tests/README
+++ b/tools/memory-model/litmus-tests/README
@@ -18,7 +18,7 @@ CoWW+poonceonce.litmus
Test of write-write coherence, that is, whether or not two
successive writes to the same variable are ordered.
-IRIW+mbonceonces+OnceOnce.litmus
+IRIW+fencembonceonces+OnceOnce.litmus
Test of independent reads from independent writes with smp_mb()
between each pairs of reads. In other words, is smp_mb()
sufficient to cause two different reading processes to agree on
@@ -47,7 +47,7 @@ ISA2+pooncerelease+poacquirerelease+poacquireonce.litmus
Can a release-acquire chain order a prior store against
a later load?
-LB+ctrlonceonce+mbonceonce.litmus
+LB+fencembonceonce+ctrlonceonce.litmus
Does a control dependency and an smp_mb() suffice for the
load-buffering litmus test, where each process reads from one
of two variables then writes to the other?
@@ -88,14 +88,14 @@ MP+porevlocks.litmus
As below, but with the first access of the writer process
and the second access of reader process protected by a lock.
-MP+wmbonceonce+rmbonceonce.litmus
+MP+fencewmbonceonce+fencermbonceonce.litmus
Does a smp_wmb() (between the stores) and an smp_rmb() (between
the loads) suffice for the message-passing litmus test, where one
process writes data and then a flag, and the other process reads
the flag and then the data. (This is similar to the ISA2 tests,
but with two processes instead of three.)
-R+mbonceonces.litmus
+R+fencembonceonces.litmus
This is the fully ordered (via smp_mb()) version of one of
the classic counterintuitive litmus tests that illustrates the
effects of store propagation delays.
@@ -103,7 +103,7 @@ R+mbonceonces.litmus
R+poonceonces.litmus
As above, but without the smp_mb() invocations.
-SB+mbonceonces.litmus
+SB+fencembonceonces.litmus
This is the fully ordered (again, via smp_mb() version of store
buffering, which forms the core of Dekker's mutual-exclusion
algorithm.
@@ -123,12 +123,12 @@ SB+rfionceonce-poonceonces.litmus
S+poonceonces.litmus
As below, but without the smp_wmb() and acquire load.
-S+wmbonceonce+poacquireonce.litmus
+S+fencewmbonceonce+poacquireonce.litmus
Can a smp_wmb(), instead of a release, and an acquire order
a prior store against a subsequent store?
WRC+poonceonces+Once.litmus
-WRC+pooncerelease+rmbonceonce+Once.litmus
+WRC+pooncerelease+fencermbonceonce+Once.litmus
These two are members of an extension of the MP litmus-test
class in which the first write is moved to a separate process.
The second is forbidden because smp_store_release() is
@@ -143,7 +143,7 @@ Z6.0+pooncelock+poonceLock+pombonce.litmus
As above, but with smp_mb__after_spinlock() immediately
following the spin_lock().
-Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus
+Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
Is the ordering provided by a release-acquire chain sufficient
to make ordering apparent to accesses by a process that does
not participate in that release-acquire chain?
diff --git a/tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus b/tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus
similarity index 90%
rename from tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus
rename to tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus
index c53350205d28..18479823cd6c 100644
--- a/tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus
+++ b/tools/memory-model/litmus-tests/S+fencewmbonceonce+poacquireonce.litmus
@@ -1,4 +1,4 @@
-C S+wmbonceonce+poacquireonce
+C S+fencewmbonceonce+poacquireonce
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/SB+mbonceonces.litmus b/tools/memory-model/litmus-tests/SB+fencembonceonces.litmus
similarity index 95%
rename from tools/memory-model/litmus-tests/SB+mbonceonces.litmus
rename to tools/memory-model/litmus-tests/SB+fencembonceonces.litmus
index 74b874ffa8da..ed5fff18d223 100644
--- a/tools/memory-model/litmus-tests/SB+mbonceonces.litmus
+++ b/tools/memory-model/litmus-tests/SB+fencembonceonces.litmus
@@ -1,4 +1,4 @@
-C SB+mbonceonces
+C SB+fencembonceonces
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus b/tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus
similarity index 93%
rename from tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus
rename to tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus
index ad3448b941e6..e9947250d7de 100644
--- a/tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus
+++ b/tools/memory-model/litmus-tests/WRC+pooncerelease+fencermbonceonce+Once.litmus
@@ -1,4 +1,4 @@
-C WRC+pooncerelease+rmbonceonce+Once
+C WRC+pooncerelease+fencermbonceonce+Once
(*
* Result: Never
diff --git a/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus b/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
similarity index 94%
rename from tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus
rename to tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
index a20fc3fafb53..88e70b87a683 100644
--- a/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus
+++ b/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+fencembonceonce.litmus
@@ -1,4 +1,4 @@
-C Z6.0+pooncerelease+poacquirerelease+mbonceonce
+C Z6.0+pooncerelease+poacquirerelease+fencembonceonce
(*
* Result: Sometimes
^ permalink raw reply related [flat|nested] 28+ messages in thread
end of thread, other threads:[~2018-07-17 9:33 UTC | newest]
Thread overview: 28+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-07-16 18:05 [PATCH memory-model 0/14] Updates to the formal memory model Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 01/14] tools/memory-model: Add litmus test for full multicopy atomicity Paul E. McKenney
2018-07-17 9:25 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 02/14] tools/memory-model: Fix ISA2+pooncelock+pooncelock+pombonce name Paul E. McKenney
2018-07-17 9:26 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 03/14] MAINTAINERS: Add Daniel Lustig as an LKMM reviewer Paul E. McKenney
2018-07-17 9:26 ` [tip:locking/core] " tip-bot for Palmer Dabbelt
2018-07-16 18:05 ` [PATCH memory-model 04/14] locking/memory-barriers.txt/kokr: Update Korean translation to fix broken DMA vs. MMIO ordering example Paul E. McKenney
2018-07-17 9:27 ` [tip:locking/core] " tip-bot for SeongJae Park
2018-07-16 18:05 ` [PATCH memory-model 05/14] tools/memory-model: Remove ACCESS_ONCE() from recipes Paul E. McKenney
2018-07-17 9:27 ` [tip:locking/core] " tip-bot for Mark Rutland
2018-07-16 18:05 ` [PATCH memory-model 06/14] tools/memory-model: Remove ACCESS_ONCE() from model Paul E. McKenney
2018-07-17 9:28 ` [tip:locking/core] " tip-bot for Mark Rutland
2018-07-16 18:05 ` [PATCH memory-model 07/14] tools/memory-model: Make scripts executable Paul E. McKenney
2018-07-17 9:29 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:05 ` [PATCH memory-model 08/14] docs: atomic_ops: Describe atomic_set as a write operation Paul E. McKenney
2018-07-16 18:06 ` [PATCH memory-model 09/14] tools/memory-model: Add informal LKMM documentation to MAINTAINERS Paul E. McKenney
2018-07-17 9:30 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-07-16 18:06 ` [PATCH memory-model 10/14] sched: Use smp_mb() in wake_woken_function() Paul E. McKenney
2018-07-17 9:30 ` [tip:locking/core] sched/core: " tip-bot for Andrea Parri
2018-07-16 18:06 ` [PATCH memory-model 11/14] locking: Clarify requirements for smp_mb__after_spinlock() Paul E. McKenney
2018-07-17 9:31 ` [tip:locking/core] locking/spinlock, sched/core: " tip-bot for Andrea Parri
2018-07-16 18:06 ` [PATCH memory-model 12/14] doc: Update wake_up() & co. memory-barrier guarantees Paul E. McKenney
2018-07-17 9:31 ` [tip:locking/core] sched/Documentation: " tip-bot for Andrea Parri
2018-07-16 18:06 ` [PATCH memory-model 13/14] memory-model/Documentation: Fix typo, smb->smp Paul E. McKenney
2018-07-17 9:32 ` [tip:locking/core] tools/memory-model/Documentation: " tip-bot for Yauheni Kaliuta
2018-07-16 18:06 ` [PATCH memory-model 14/14] tools/memory-model: Rename litmus tests to comply to norm7 Paul E. McKenney
2018-07-17 9:33 ` [tip:locking/core] " tip-bot for Andrea Parri
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).