Miscellaneous fixes
Subsystems (2)
RFC PATCH 1/1 Miscellaneous fixes
Hello!
This RFC series adds some miscellaneous updates to the Linux kernel memory model:
1. Clarify the origin and scope of the tool name to avoid confusion
between "memory model" and "memory management", courtesy of Andrea Parri.
2. Move the maintainer list for LKMM to the main MAINTAINERS file,
courtesy of Andrea Parri.
3. Add memory-barriers.txt to the LKMM MAINTAINERS entry, courtesy
of Andrea Parri.
4. Add comments explaining the purpose of the various litmus tests.
5. Fix puntuation errors in litmus-tests/README.
6. Add LKMM acronym to MAINTAINERS entry. If there are no objections,
this will be squashed with #3 above.
7. Add Akira Yokosawa as an LKMM reviewer.
8-10. Replace underscores with hyphens for smp_mb__*() workings. Note that
this patch leaves rb_dep alone because the underscore is set by herd7. If there are no objections, these will be squashed together.
Thanx, Paul
------------------------------------------------------------------------
/tools/memory-model/MAINTAINERS | 15 ------- b/MAINTAINERS | 20 +++++++++- b/tools/memory-model/MAINTAINERS | 2 - b/tools/memory-model/README | 14 +++---- b/tools/memory-model/linux-kernel.bell | 8 ++-- b/tools/memory-model/linux-kernel.cat | 8 ++-- b/tools/memory-model/linux-kernel.def | 6 +-- b/tools/memory-model/litmus-tests/CoRR+poonceonce+Once.litmus | 7 +++ b/tools/memory-model/litmus-tests/CoRW+poonceonce+Once.litmus | 7 +++ b/tools/memory-model/litmus-tests/CoWR+poonceonce+Once.litmus | 7 +++ b/tools/memory-model/litmus-tests/CoWW+poonceonce.litmus | 7 +++ b/tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus | 10 +++++ b/tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus | 10 +++++ b/tools/memory-model/litmus-tests/ISA2+poonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/ISA2+pooncerelease+poacquirerelease+poacquireonce.litmus | 11 +++++ b/tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus | 11 +++++ b/tools/memory-model/litmus-tests/LB+poacquireonce+pooncerelease.litmus | 8 ++++ b/tools/memory-model/litmus-tests/LB+poonceonces.litmus | 7 +++ b/tools/memory-model/litmus-tests/MP+onceassign+derefonce.litmus | 11 +++++ b/tools/memory-model/litmus-tests/MP+polocks.litmus | 11 +++++ b/tools/memory-model/litmus-tests/MP+poonceonces.litmus | 7 +++ b/tools/memory-model/litmus-tests/MP+pooncerelease+poacquireonce.litmus | 8 ++++ b/tools/memory-model/litmus-tests/MP+porevlocks.litmus | 11 +++++ b/tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus | 8 ++++ b/tools/memory-model/litmus-tests/R+mbonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/R+poonceonces.litmus | 8 ++++ b/tools/memory-model/litmus-tests/README | 4 +- b/tools/memory-model/litmus-tests/S+poonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus | 7 +++ b/tools/memory-model/litmus-tests/SB+mbonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/SB+poonceonces.litmus | 8 ++++ b/tools/memory-model/litmus-tests/WRC+poonceonces+Once.litmus | 8 ++++ b/tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus | 8 ++++ b/tools/memory-model/litmus-tests/Z6.0+pooncelock+poonceLock+pombonce.litmus | 9 ++++ b/tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus | 8 ++++ b/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus | 14 +++++++ 36 files changed, 286 insertions(+), 38 deletions(-)
RFC PATCH 1/10 tools/memory-model: Clarify the origin/scope of the tool name
From: Andrea Parri <[email protected]>
Ingo pointed out that:
"The "memory model" name is overly generic, ambiguous and somewhat
misleading, as we usually mean the virtual memory layout/model when we say "memory model". GCC too uses it in that sense [...]"
Make it clear that tools/memory-model/ uses the term "memory model" as shorthand for "memory consistency model" by calling out this convention in tools/memory-model/README.
Stick to the original "memory model" term in sources' headers and for the subsystem name.
--- tools/memory-model/MAINTAINERS | 2 +- tools/memory-model/README | 14 +++++++------- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- 4 files changed, 10 insertions(+), 10 deletions(-)
tools/memory-model/MAINTAINERS +1 -1
@@ -1,4 +1,4 @@ -LINUX KERNEL MEMORY MODEL +LINUX KERNEL MEMORY CONSISTENCY MODEL M: Alan Stern <[email protected]> M: Andrea Parri <[email protected]> M: Will Deacon <[email protected]>
tools/memory-model/README +7 -7
@@ -1,15 +1,15 @@ - ========================= - LINUX KERNEL MEMORY MODEL - ========================= + ===================================== + LINUX KERNEL MEMORY CONSISTENCY MODEL + ===================================== ============ INTRODUCTION ============ -This directory contains the memory model of the Linux kernel, written -in the "cat" language and executable by the (externally provided) -"herd7" simulator, which exhaustively explores the state space of -small litmus tests. +This directory contains the memory consistency model (memory model, for +short) of the Linux kernel, written in the "cat" language and executable +by the externally provided "herd7" simulator, which exhaustively explores +the state space of small litmus tests. In addition, the "klitmus7" tool (also externally provided) may be used to convert a litmus test to a Linux kernel module, which in turn allows
tools/memory-model/linux-kernel.bell +1 -1
@@ -11,7 +11,7 @@ * which is to appear in ASPLOS 2018. *) -"Linux kernel memory model" +"Linux-kernel memory consistency model" enum Accesses = 'once (*READ_ONCE,WRITE_ONCE,ACCESS_ONCE*) || 'release (*smp_store_release*) ||
tools/memory-model/linux-kernel.cat +1 -1
@@ -11,7 +11,7 @@ * which is to appear in ASPLOS 2018. *) -"Linux kernel memory model" +"Linux-kernel memory consistency model" (* * File "lock.cat" handles locks and is experimental.
Signature · 1 line
2.5.2
RFC PATCH 2/10 MAINTAINERS: Add the Memory Consistency Model subsystem
In reply to Paul E. McKenneyFrom: Andrea Parri <[email protected]>
Move the contents of tools/memory-model/MAINTAINERS into the main MAINTAINERS file, removing tools/memory-model/MAINTAINERS. This allows get_maintainer.pl to correctly identify the maintainers of tools/memory-model/.
--- MAINTAINERS | 16 ++++++++++++++++ tools/memory-model/MAINTAINERS | 15 --------------- 2 files changed, 16 insertions(+), 15 deletions(-) delete mode 100644 tools/memory-model/MAINTAINERS
MAINTAINERS +16 -0
@@ -8086,6 +8086,22 @@ M: Kees Cook <[email protected]> S: Maintained F: drivers/misc/lkdtm* +LINUX KERNEL MEMORY CONSISTENCY MODEL +M: Alan Stern <[email protected]> +M: Andrea Parri <[email protected]> +M: Will Deacon <[email protected]> +M: Peter Zijlstra <[email protected]> +M: Boqun Feng <[email protected]> +M: Nicholas Piggin <[email protected]> +M: David Howells <[email protected]> +M: Jade Alglave <[email protected]> +M: Luc Maranget <[email protected]> +M: "Paul E. McKenney" <[email protected]> +L: [email protected] +S: Supported +T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git +F: tools/memory-model/ + LINUX SECURITY MODULE (LSM) FRAMEWORK M: Chris Wright <[email protected]> L: [email protected]
tools/memory-model/MAINTAINERS deleted +0 -15
@@ -1,15 +0,0 @@ -LINUX KERNEL MEMORY CONSISTENCY MODEL -M: Alan Stern <[email protected]> -M: Andrea Parri <[email protected]> -M: Will Deacon <[email protected]> -M: Peter Zijlstra <[email protected]> -M: Boqun Feng <[email protected]> -M: Nicholas Piggin <[email protected]> -M: David Howells <[email protected]> -M: Jade Alglave <[email protected]> -M: Luc Maranget <[email protected]> -M: "Paul E. McKenney" <[email protected]> -L: [email protected] -S: Supported -T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git -F: tools/memory-model/
Signature · 1 line
2.5.2
RFC PATCH 3/10 MAINTAINERS: List file memory-barriers.txt within the LKMM entry
In reply to Paul E. McKenneyFrom: Andrea Parri <[email protected]>
We now have a shiny new Linux-kernel memory model (LKMM) and the old tried-and-true Documentation/memory-barrier.txt. It would be good to keep these automatically synchronized, but in the meantime we need at least let people know that they are related. Will suggested adding the Documentation/memory-barrier.txt file to the LKMM maintainership list, thus making the LKMM maintainers responsible for both the old and the new. This commit follows Will's excellent suggestion.
--- MAINTAINERS | 1 + 1 file changed, 1 insertion(+)
MAINTAINERS +1 -0
@@ -8101,6 +8101,7 @@ L: [email protected] S: Supported T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git F: tools/memory-model/ +F: Documentation/memory-barriers.txt LINUX SECURITY MODULE (LSM) FRAMEWORK M: Chris Wright <[email protected]>
Signature · 1 line
2.5.2
RFC PATCH 4/10 EXP litmus_tests: Add comments explaining tests' purposes
In reply to Paul E. McKenneyThis commit adds comments to the litmus tests summarizing what these tests are intended to demonstrate.
--- .../memory-model/litmus-tests/CoRR+poonceonce+Once.litmus | 7 +++++++ .../memory-model/litmus-tests/CoRW+poonceonce+Once.litmus | 7 +++++++ .../memory-model/litmus-tests/CoWR+poonceonce+Once.litmus | 7 +++++++ tools/memory-model/litmus-tests/CoWW+poonceonce.litmus | 7 +++++++ .../litmus-tests/IRIW+mbonceonces+OnceOnce.litmus | 10 ++++++++++ .../litmus-tests/IRIW+poonceonces+OnceOnce.litmus | 10 ++++++++++ tools/memory-model/litmus-tests/ISA2+poonceonces.litmus | 9 +++++++++ ...SA2+pooncerelease+poacquirerelease+poacquireonce.litmus | 11 +++++++++++ .../litmus-tests/LB+ctrlonceonce+mbonceonce.litmus | 11 +++++++++++ .../litmus-tests/LB+poacquireonce+pooncerelease.litmus | 8 ++++++++ tools/memory-model/litmus-tests/LB+poonceonces.litmus | 7 +++++++ .../litmus-tests/MP+onceassign+derefonce.litmus | 11 ++++++++++- tools/memory-model/litmus-tests/MP+polocks.litmus | 11 +++++++++++ tools/memory-model/litmus-tests/MP+poonceonces.litmus | 7 +++++++ .../litmus-tests/MP+pooncerelease+poacquireonce.litmus | 8 ++++++++ tools/memory-model/litmus-tests/MP+porevlocks.litmus | 11 +++++++++++ .../litmus-tests/MP+wmbonceonce+rmbonceonce.litmus | 8 ++++++++ tools/memory-model/litmus-tests/R+mbonceonces.litmus | 9 +++++++++ tools/memory-model/litmus-tests/R+poonceonces.litmus | 8 ++++++++ tools/memory-model/litmus-tests/S+poonceonces.litmus | 9 +++++++++ .../litmus-tests/S+wmbonceonce+poacquireonce.litmus | 7 +++++++ tools/memory-model/litmus-tests/SB+mbonceonces.litmus | 9 +++++++++ tools/memory-model/litmus-tests/SB+poonceonces.litmus | 8 ++++++++ .../memory-model/litmus-tests/WRC+poonceonces+Once.litmus | 8 ++++++++ .../litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus | 8 ++++++++ .../Z6.0+pooncelock+poonceLock+pombonce.litmus | 9 +++++++++ .../Z6.0+pooncelock+pooncelock+pombonce.litmus | 8 ++++++++ .../Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus | 14 ++++++++++++++ 28 files changed, 246 insertions(+), 1 deletion(-)
tools/memory-model/litmus-tests/CoRR+poonceonce+Once.litmus +7 -0
@@ -1,5 +1,12 @@ C CoRR+poonceonce+Once +(* + * Result: Never + * + * Test of read-read coherence, that is, whether or not two successive + * reads from the same variable are ordered. + *) + {} P0(int *x)
tools/memory-model/litmus-tests/CoRW+poonceonce+Once.litmus +7 -0
@@ -1,5 +1,12 @@ C CoRW+poonceonce+Once +(* + * Result: Never + * + * Test of read-write coherence, that is, whether or not a read from + * a given variable and a later write to that same variable are ordered. + *) + {} P0(int *x)
tools/memory-model/litmus-tests/CoWR+poonceonce+Once.litmus +7 -0
@@ -1,5 +1,12 @@ C CoWR+poonceonce+Once +(* + * Result: Never + * + * Test of write-read coherence, that is, whether or not a write to a + * given variable and a later read from that same variable are ordered. + *) + {} P0(int *x)
tools/memory-model/litmus-tests/CoWW+poonceonce.litmus +7 -0
@@ -1,5 +1,12 @@ C CoWW+poonceonce +(* + * Result: Never + * + * Test of write-write coherence, that is, whether or not two successive + * writes to the same variable are ordered. + *) + {} P0(int *x)
tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus +10 -0
@@ -1,5 +1,15 @@ C IRIW+mbonceonces+OnceOnce +(* + * Result: Never + * + * 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 the order of a pair + * of writes, where each write is to a different variable by a different + * process? + *) + {} P0(int *x)
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus +10 -0
@@ -1,5 +1,15 @@ C IRIW+poonceonces+OnceOnce +(* + * Result: Never + * + * Test of independent reads from independent writes with nothing + * between each pairs of reads. In other words, is anything at all + * needed to cause two different reading processes to agree on the order + * of a pair of writes, where each write is to a different variable by a + * different process? + *) + {} P0(int *x)
tools/memory-model/litmus-tests/ISA2+poonceonces.litmus +9 -0
@@ -1,5 +1,14 @@ C ISA2+poonceonces +(* + * Result: Sometimes + * + * Given a release-acquire chain ordering the first process's store + * against the last process's load, is ordering preserved if all of the + * smp_store_release() invocations are replaced by WRITE_ONCE() and all + * of the smp_load_acquire() invocations are replaced by READ_ONCE()? + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/ISA2+pooncerelease+poacquirerelease+poacquireonce.litmus +11 -0
@@ -1,5 +1,16 @@ C ISA2+pooncerelease+poacquirerelease+poacquireonce +(* + * Result: Never + * + * This litmus test demonstrates that a release-acquire chain suffices + * to order P0()'s initial write against P2()'s final read. The reason + * that the release-acquire chain suffices is because in all but one + * case (P2() to P0()), each process reads from the preceding process's + * write. In memory-model-speak, there is only one non-reads-from + * (AKA non-rf) link, so release-acquire is all that is needed. + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus +11 -0
@@ -1,5 +1,16 @@ C LB+ctrlonceonce+mbonceonce +(* + * Result: Never + * + * This litmus test demonstrates that lightweight ordering suffices for + * the load-buffering pattern, in other words, preventing all processes + * reading from the preceding process's write. In this example, the + * combination of a control dependency and a full memory barrier are enough + * to do the trick. (But the full memory barrier could be replaced with + * another control dependency and order would still be maintained.) + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/LB+poacquireonce+pooncerelease.litmus +8 -0
@@ -1,5 +1,13 @@ C LB+poacquireonce+pooncerelease +(* + * Result: Never + * + * Does a release-acquire pair suffice for the load-buffering litmus + * test, where each process reads from one of two variables then writes + * to the other? + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/LB+poonceonces.litmus +7 -0
@@ -1,5 +1,12 @@ C LB+poonceonces +(* + * Result: Sometimes + * + * Can the counter-intuitive outcome for the load-buffering pattern + * be prevented even with no explicit ordering? + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/MP+onceassign+derefonce.litmus +10 -1
@@ -1,4 +1,13 @@ -C MP+onceassign+derefonce.litmus +C MP+onceassign+derefonce + +(* + * Result: Never + * + * This litmus test demonstrates that rcu_assign_pointer() and + * rcu_dereference() suffice to ensure that an RCU reader will not see + * pre-initialization garbage when it traverses an RCU-protected data + * structure containing a newly inserted element. + *) { y=z;
tools/memory-model/litmus-tests/MP+polocks.litmus +11 -0
@@ -1,5 +1,16 @@ C MP+polocks +(* + * Result: Never + * + * This litmus test demonstrates how lock acquisitions and releases can + * stand in for smp_load_acquire() and smp_store_release(), respectively. + * In other words, when holding a given lock (or indeed after releasing a + * given lock), a CPU is not only guaranteed to see the accesses that other + * CPUs made while previously holding that lock, it is also guaranteed + * to see all prior accesses by those other CPUs. + *) + {} P0(int *x, int *y, spinlock_t *mylock)
tools/memory-model/litmus-tests/MP+poonceonces.litmus +7 -0
@@ -1,5 +1,12 @@ C MP+poonceonces +(* + * Result: Maybe + * + * Can the counter-intuitive message-passing outcome be prevented with + * no ordering at all? + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/MP+pooncerelease+poacquireonce.litmus +8 -0
@@ -1,5 +1,13 @@ C MP+pooncerelease+poacquireonce +(* + * Result: Never + * + * This litmus test demonstrates that smp_store_release() and + * smp_load_acquire() provide sufficient ordering for the message-passing + * pattern. + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/MP+porevlocks.litmus +11 -0
@@ -1,5 +1,16 @@ C MP+porevlocks +(* + * Result: Never + * + * This litmus test demonstrates how lock acquisitions and releases can + * stand in for smp_load_acquire() and smp_store_release(), respectively. + * In other words, when holding a given lock (or indeed after releasing a + * given lock), a CPU is not only guaranteed to see the accesses that other + * CPUs made while previously holding that lock, it is also guaranteed to + * see all prior accesses by those other CPUs. + *) + {} P0(int *x, int *y, spinlock_t *mylock)
tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus +8 -0
@@ -1,5 +1,13 @@ C MP+wmbonceonce+rmbonceonce +(* + * Result: Never + * + * This litmus test demonstrates that smp_wmb() and smp_rmb() provide + * sufficient ordering for the message-passing pattern. However, it + * is usually better to use smp_store_release() and smp_load_acquire(). + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/R+mbonceonces.litmus +9 -0
@@ -1,5 +1,14 @@ C R+mbonceonces +(* + * Result: Never + * + * 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. Note that weakening either of the barriers would + * cause the resulting test to be allowed. + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/R+poonceonces.litmus +8 -0
@@ -1,5 +1,13 @@ C R+poonceonces +(* + * Result: Sometimes + * + * This is the unordered (thus lacking smp_mb()) version of one of the + * classic counterintuitive litmus tests that illustrates the effects of + * store propagation delays. + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/S+poonceonces.litmus +9 -0
@@ -1,5 +1,14 @@ C S+poonceonces +(* + * Result: Sometimes + * + * Starting with a two-process release-acquire chain ordering P0()'s + * first store against P1()'s final load, if the smp_store_release() + * is replaced by WRITE_ONCE() and the smp_load_acquire() replaced by + * READ_ONCE(), is ordering preserved? + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus +7 -0
@@ -1,5 +1,12 @@ C S+wmbonceonce+poacquireonce +(* + * Result: Never + * + * Can a smp_wmb(), instead of a release, and an acquire order a prior + * store against a subsequent store? + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/SB+mbonceonces.litmus +9 -0
@@ -1,5 +1,14 @@ C SB+mbonceonces +(* + * Result: Never + * + * This litmus test demonstrates that full memory barriers suffice to + * order the store-buffering pattern, where each process writes to the + * variable that the preceding process reads. (Locking and RCU can also + * suffice, but not much else.) + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/SB+poonceonces.litmus +8 -0
@@ -1,5 +1,13 @@ C SB+poonceonces +(* + * Result: Sometimes + * + * This litmus test demonstrates that at least some ordering is required + * to order the store-buffering pattern, where each process writes to the + * variable that the preceding process reads. + *) + {} P0(int *x, int *y)
tools/memory-model/litmus-tests/WRC+poonceonces+Once.litmus +8 -0
@@ -1,5 +1,13 @@ C WRC+poonceonces+Once +(* + * Result: Sometimes + * + * This litmus test is an extension of the message-passing pattern, + * where the first write is moved to a separate process. Note that this + * test has no ordering at all. + *) + {} P0(int *x)
tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus +8 -0
@@ -1,5 +1,13 @@ C WRC+pooncerelease+rmbonceonce+Once +(* + * Result: Never + * + * This litmus test is an extension of the message-passing pattern, where + * the first write is moved to a separate process. Because it features + * a release and a read memory barrier, it should be forbidden. + *) + {} P0(int *x)
tools/memory-model/litmus-tests/Z6.0+pooncelock+poonceLock+pombonce.litmus +9 -0
@@ -1,5 +1,14 @@ C Z6.0+pooncelock+poonceLock+pombonce +(* + * Result: Never + * + * This litmus test demonstrates how smp_mb__after_spinlock() may be + * used to ensure that accesses in different critical sections for a + * given lock running on different CPUs are nevertheless seen in order + * by CPUs not holding that lock. + *) + {} P0(int *x, int *y, spinlock_t *mylock)
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +8 -0
@@ -1,5 +1,13 @@ C Z6.0+pooncelock+pooncelock+pombonce +(* + * Result: Never + * + * This example demonstrates that a pair of accesses made by different + * processes each while holding a given lock will not necessarily be + * seen as ordered by a third process not holding that lock. + *) + {} P0(int *x, int *y, spinlock_t *mylock)
tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus +14 -0
@@ -1,5 +1,19 @@ C Z6.0+pooncerelease+poacquirerelease+mbonceonce +(* + * Result: Sometimes + * + * This litmus test shows that a release-acquire chain, while sufficient + * when there is but one non-reads-from (AKA non-rf) link, does not suffice + * if there is more than one. Of the three processes, only P1() reads from + * P0's write, which means that there are two non-rf links: P1() to P2() + * is a write-to-write link (AKA a "coherence" or just "co" link) and P2() + * to P0() is a read-to-write link (AKA a "from-reads" or just "fr" link). + * When there are two or more non-rf links, you typically will need one + * full barrier for each non-rf link. (Exceptions include some cases + * involving locking.) + *) + {} P0(int *x, int *y)
Signature · 1 line
2.5.2
Alan Stern
maintainerRe: [PATCH RFC tip/lkmm 04/10] EXP litmus_tests: Add comments explaining tests' purposes
Quoted 12 lines · Paul E. McKenney
This commit adds comments to the litmus tests summarizing what these tests are intended to demonstrate.
Suggested-by Ingo Molnar <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>[ paulmck: Apply Andrea's and Alan's feedback. ]---
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus +0
-0@@ -1,5 +1,15 @@ C IRIW+poonceonces+OnceOnce
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus line 11
Quoted 10 lines
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus +9
-0+(* + * Result: Never + * + * Test of independent reads from independent writes with nothing + * between each pairs of reads. In other words, is anything at all + * needed to cause two different reading processes to agree on the order + * of a pair of writes, where each write is to a different variable by a + * different process? + *)
The result should be Sometimes.
Alan
Re: [PATCH RFC tip/lkmm 04/10] EXP litmus_tests: Add comments explaining tests' purposes
Quoted 15 lines · Alan Stern
Quoted 13 lines · Paul E. McKenney
This commit adds comments to the litmus tests summarizing what these tests are intended to demonstrate.
Suggested-by Ingo Molnar <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>[ paulmck: Apply Andrea's and Alan's feedback. ]---
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus +0
-0@@ -1,5 +1,15 @@ C IRIW+poonceonces+OnceOnce
Quoted 11 lines
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus line 11
Quoted 9 lines
tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus +9
-0+(* + * Result: Never + * + * Test of independent reads from independent writes with nothing + * between each pairs of reads. In other words, is anything at all + * needed to cause two different reading processes to agree on the order + * of a pair of writes, where each write is to a different variable by a + * different process? + *)The result should be Sometimes.
Right you are, and better yet, the memory model agrees with you.
I have fixed this, thank you for the careful review!
Thanx, Paul
RFC PATCH 5/10 README: Fix a couple of punctuation errors
In reply to Paul E. McKenney--- tools/memory-model/litmus-tests/README | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tools/memory-model/litmus-tests/README +2 -2
@@ -23,14 +23,14 @@ IRIW+mbonceonces+OnceOnce.litmus between each pairs of reads. In other words, is smp_mb() sufficient to cause two different reading processes to agree on the order of a pair of writes, where each write is to a different - variable by a different process. + variable by a different process? IRIW+poonceonces+OnceOnce.litmus Test of independent reads from independent writes with nothing between each pairs of reads. In other words, is anything at all needed to cause two different reading processes to agree on the order of a pair of writes, where each write is to a different - variable by a different process. + variable by a different process? ISA2+poonceonces.litmus As below, but with store-release replaced with WRITE_ONCE()
Signature · 1 line
2.5.2
RFC PATCH 6/10 EXP MAINTAINERS: Add the "LKMM" acronym
In reply to Paul E. McKenneyThis commit adds "(LKMM)" to the end of the MAINTAINERS file's entry for the Linux-kernel memory-consistency model.
If people are OK with this, it should be squashed into 12a62a1d0703 ("MAINTAINERS: Add the Memory Consistency Model subsystem").
--- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
MAINTAINERS +1 -1
@@ -8086,7 +8086,7 @@ M: Kees Cook <[email protected]> S: Maintained F: drivers/misc/lkdtm* -LINUX KERNEL MEMORY CONSISTENCY MODEL +LINUX KERNEL MEMORY CONSISTENCY MODEL (LKMM) M: Alan Stern <[email protected]> M: Andrea Parri <[email protected]> M: Will Deacon <[email protected]>
Signature · 1 line
2.5.2
RFC PATCH 7/10 MAINTAINERS: Add Akira Yokosawa as an LKMM reviewer
In reply to Paul E. McKenney--- MAINTAINERS | 1 + 1 file changed, 1 insertion(+)
MAINTAINERS +1 -0
@@ -8097,6 +8097,7 @@ M: David Howells <[email protected]> M: Jade Alglave <[email protected]> M: Luc Maranget <[email protected]> M: "Paul E. McKenney" <[email protected]> +R: Akira Yokosawa <[email protected]> L: [email protected] S: Supported T: git git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
Signature · 1 line
2.5.2
RFC PATCH 8/10 EXP Remove understore from smp_mb__before_atomic() workings
In reply to Paul E. McKenney--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)
tools/memory-model/linux-kernel.bell +1 -1
@@ -28,7 +28,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) || - 'before_atomic (*smp_mb__before_atomic*) || + 'before-atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers]
tools/memory-model/linux-kernel.cat +1 -1
@@ -29,7 +29,7 @@ let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | - ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | + ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po?
tools/memory-model/linux-kernel.def +1 -1
@@ -21,7 +21,7 @@ smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } -smp_mb__before_atomic() { __fence{before_atomic} ; } +smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after_atomic} ; } smp_mb__after_spinlock() { __fence{after_spinlock} ; }
Signature · 1 line
2.5.2
Andrea Parri
maintainerRe: [PATCH RFC tip/lkmm 08/10] EXP Remove understore from smp_mb__before_atomic() workings
Quoted 1 line · Paul E. McKenney
Signed-off-by Paul E. McKenney <[email protected]>
nit: s/understore/underscore (a very understandable lapsus, BTW ;-)
According to the cover letter, patches 8-10 of this series are intended to be squashed together; it could be useful to add a description then:
the motivation for these changes might appear "trivial" today; months/ years from now? mmh, I can't put my finger on it ;-)
IMO, this descrition should also mention why this solution "is the way to go" (e.g., compare with the solution recently implemented by Akira).
Andrea
Quoted 48 lines
--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)tools/memory-model/linux-kernel.bell +1
-1@@ -28,7 +28,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) || - 'before_atomic (*smp_mb__before_atomic*) || + 'before-atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers]tools/memory-model/linux-kernel.cat +1
-1@@ -29,7 +29,7 @@ let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | - ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | + ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po?tools/memory-model/linux-kernel.def +1
-1@@ -21,7 +21,7 @@ smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } -smp_mb__before_atomic() { __fence{before_atomic} ; } +smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after_atomic} ; } smp_mb__after_spinlock() { __fence{after_spinlock} ; }Signature · 1 line
2.5.2
Re: [PATCH RFC tip/lkmm 08/10] EXP Remove understore from smp_mb__before_atomic() workings
Quoted 4 lines · Andrea Parri
Quoted 1 line · Paul E. McKenney
Signed-off-by Paul E. McKenney <[email protected]>nit: s/understore/underscore (a very understandable lapsus, BTW ;-)
;-)
Will fix when squashing.
Quoted 4 lines
According to the cover letter, patches 8-10 of this series are intended to be squashed together; it could be useful to add a description then:
the motivation for these changes might appear "trivial" today; months/ years from now? mmh, I can't put my finger on it ;-)
Assuming we keep these patches, agreed. Something like this for the aggregated single commit, perhaps:
tools/memory-model: Convert underscores to hyphens
Typical cat-language code uses hyphens for word separators in identifiers, but several LKMM identifiers use underscores instead. This commit therefore converts the underscores in the .bell and .cat files for smp_mb__before_atomic(), smp_mb__after_atomic(), and smp_mb__after_spinlock() to hyphens.
Quoted 2 lines
IMO, this descrition should also mention why this solution "is the way to go" (e.g., compare with the solution recently implemented by Akira).
You lost me here. Or are you asking for a commit log similar to that provided by Akira for his forward-port of Alan's unrelated patch? If so, please see above.
Thanx, Paul
Quoted 52 lines
Andrea
Quoted 48 lines
--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)tools/memory-model/linux-kernel.bell +1
-1@@ -28,7 +28,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) || - 'before_atomic (*smp_mb__before_atomic*) || + 'before-atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers]tools/memory-model/linux-kernel.cat +1
-1@@ -29,7 +29,7 @@ let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | - ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | + ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po?tools/memory-model/linux-kernel.def +1
-1@@ -21,7 +21,7 @@ smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } -smp_mb__before_atomic() { __fence{before_atomic} ; } +smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after_atomic} ; } smp_mb__after_spinlock() { __fence{after_spinlock} ; }Signature · 1 line
2.5.2
Andrea Parri
maintainerRe: [PATCH RFC tip/lkmm 08/10] EXP Remove understore from smp_mb__before_atomic() workings
Quoted 26 lines · Paul E. McKenney
Quoted 4 lines · Andrea Parri
Quoted 1 line · Paul E. McKenney
Signed-off-by Paul E. McKenney <[email protected]>nit: s/understore/underscore (a very understandable lapsus, BTW ;-)
;-)
Will fix when squashing.
Quoted 4 lines
According to the cover letter, patches 8-10 of this series are intended to be squashed together; it could be useful to add a description then:
the motivation for these changes might appear "trivial" today; months/ years from now? mmh, I can't put my finger on it ;-)Assuming we keep these patches, agreed. Something like this for the aggregated single commit, perhaps:
tools/memory-model: Convert underscores to hyphensTypical cat-language code uses hyphens for word separators in identifiers, but several LKMM identifiers use underscores instead. This commit therefore converts the underscores in the .bell and .cat files for
smp_mb__before_atomic(),smp_mb__after_atomic(), andsmp_mb__after_spinlock()to hyphens.
Quoted 6 lines
Quoted 2 lines
IMO, this descrition should also mention why this solution "is the way to go" (e.g., compare with the solution recently implemented by Akira).
You lost me here. Or are you asking for a commit log similar to that provided by Akira for his forward-port of Alan's unrelated patch? If so, please see above.
I was mainly thinking "but rb_dep..."; or you're also planning to rename this annotation and, maybe, the "rb-dep" relation (in order to avoid the conflict)?
Andrea
Quoted 56 lines
Thanx, Paul
Quoted 52 lines
Andrea
Quoted 48 lines
--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)tools/memory-model/linux-kernel.bell +1
-1@@ -28,7 +28,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) || - 'before_atomic (*smp_mb__before_atomic*) || + 'before-atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers]tools/memory-model/linux-kernel.cat +1
-1@@ -29,7 +29,7 @@ let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | - ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | + ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po?tools/memory-model/linux-kernel.def +1
-1@@ -21,7 +21,7 @@ smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } -smp_mb__before_atomic() { __fence{before_atomic} ; } +smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after_atomic} ; } smp_mb__after_spinlock() { __fence{after_spinlock} ; }Signature · 1 line
2.5.2
Re: [PATCH RFC tip/lkmm 08/10] EXP Remove understore from smp_mb__before_atomic() workings
Quoted 30 lines · Andrea Parri
Quoted 29 lines · Paul E. McKenney
Quoted 4 lines · Andrea Parri
Quoted 1 line · Paul E. McKenney
Signed-off-by Paul E. McKenney <[email protected]>nit: s/understore/underscore (a very understandable lapsus, BTW ;-)
;-)
Will fix when squashing.
Quoted 4 lines
According to the cover letter, patches 8-10 of this series are intended to be squashed together; it could be useful to add a description then:
the motivation for these changes might appear "trivial" today; months/ years from now? mmh, I can't put my finger on it ;-)Assuming we keep these patches, agreed. Something like this for the aggregated single commit, perhaps:
tools/memory-model: Convert underscores to hyphensTypical cat-language code uses hyphens for word separators in identifiers, but several LKMM identifiers use underscores instead. This commit therefore converts the underscores in the .bell and .cat files for
smp_mb__before_atomic(),smp_mb__after_atomic(), andsmp_mb__after_spinlock()to hyphens.Quoted 2 lines
IMO, this descrition should also mention why this solution "is the way to go" (e.g., compare with the solution recently implemented by Akira).
Quoted 7 lines
Quoted 3 lines
You lost me here. Or are you asking for a commit log similar to that provided by Akira for his forward-port of Alan's unrelated patch? If so, please see above.
I was mainly thinking "but
rb_dep..."; or you're also planning to rename this annotation and, maybe, the "rb-dep" relation (in order to avoid the conflict)?
The rb_dep relation goes away completely once we catch LKMM up with the recent changes that remove smp_read_barrier_depends() from the core kernel. So my current thought is hold back the "_"->"-" commit until that catch-up happens.
Thanx, Paul
Quoted 60 lines
Andrea
Quoted 56 lines
Thanx, Paul
Quoted 52 lines
Andrea
Quoted 48 lines
--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)tools/memory-model/linux-kernel.bell +1
-1@@ -28,7 +28,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) || - 'before_atomic (*smp_mb__before_atomic*) || + 'before-atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers]tools/memory-model/linux-kernel.cat +1
-1@@ -29,7 +29,7 @@ let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | - ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | + ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po?tools/memory-model/linux-kernel.def +1
-1@@ -21,7 +21,7 @@ smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } -smp_mb__before_atomic() { __fence{before_atomic} ; } +smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after_atomic} ; } smp_mb__after_spinlock() { __fence{after_spinlock} ; }Signature · 1 line
2.5.2
RFC PATCH 9/10 EXP Remove underscore from smp_mb__after_atomic() workings
In reply to Paul E. McKenney--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)
tools/memory-model/linux-kernel.bell +1 -1
@@ -29,7 +29,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) || 'before-atomic (*smp_mb__before_atomic*) || - 'after_atomic (*smp_mb__after_atomic*) || + 'after-atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers]
tools/memory-model/linux-kernel.cat +1 -1
@@ -30,7 +30,7 @@ let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | - ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | + ([M] ; po? ; [RMW] ; fencerel(After-atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po?
tools/memory-model/linux-kernel.def +1 -1
@@ -22,7 +22,7 @@ smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } -smp_mb__after_atomic() { __fence{after_atomic} ; } +smp_mb__after_atomic() { __fence{after-atomic} ; } smp_mb__after_spinlock() { __fence{after_spinlock} ; } // Exchange
Signature · 1 line
2.5.2
RFC PATCH 10/10 EXP Remove underscores from smp_mb__after_spinlock() workings
In reply to Paul E. McKenney--- tools/memory-model/linux-kernel.bell | 2 +- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)
tools/memory-model/linux-kernel.bell +1 -1
@@ -30,7 +30,7 @@ enum Barriers = 'wmb (*smp_wmb*) || 'sync-rcu (*synchronize_rcu*) || 'before-atomic (*smp_mb__before_atomic*) || 'after-atomic (*smp_mb__after_atomic*) || - 'after_spinlock (*smp_mb__after_spinlock*) + 'after-spinlock (*smp_mb__after_spinlock*) instructions F[Barriers] (* Compute matching pairs of nested Rcu-lock and Rcu-unlock *)
tools/memory-model/linux-kernel.cat +1 -1
@@ -31,7 +31,7 @@ let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After-atomic) ; [M]) | - ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) + ([M] ; po? ; [LKW] ; fencerel(After-spinlock) ; [M]) let gp = po ; [Sync-rcu] ; po? let strong-fence = mb | gp
tools/memory-model/linux-kernel.def +1 -1
@@ -23,7 +23,7 @@ smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after-atomic} ; } -smp_mb__after_spinlock() { __fence{after_spinlock} ; } +smp_mb__after_spinlock() { __fence{after-spinlock} ; } // Exchange xchg(X,V) __xchg{mb}(X,V)
Signature · 1 line
2.5.2
Akira Yokosawa
reviewerQuoted 19 lines · Paul E. McKenney
Hello!
This RFC series adds some miscellaneous updates to the Linux kernel memory model:
1. Clarify the origin and scope of the tool name to avoid confusion
between "memory model" and "memory management", courtesy of Andrea Parri.2. Move the maintainer list for LKMM to the main
MAINTAINERSfile,
courtesy of Andrea Parri.3. Add memory-barriers.txt to the LKMM
MAINTAINERSentry, courtesy
of Andrea Parri.4. Add comments explaining the purpose of the various litmus tests.
5. Fix puntuation errors in litmus-tests/README.
Quoted 8 lines
6. Add LKMM acronym to
MAINTAINERSentry. If there are no objections,
this will be squashed with #3 above.7. Add Akira Yokosawa as an LKMM reviewer.
8-10. Replace underscores with hyphens for smp_mb__*() workings. Note that
this patch leavesrb_depalone because the underscore is set by herd7. If there are no objections, these will be squashed together.
Paul, Patches 8-10 made the problem even worse with herd7 7.47. I'll submit an alternative patch in reply to this message.
Thanks, Akira
Quoted 43 lines
Thanx, Paul
------------------------------------------------------------------------
/tools/memory-model/MAINTAINERS | 15 ------- b/MAINTAINERS | 20 +++++++++- b/tools/memory-model/MAINTAINERS | 2 - b/tools/memory-model/README | 14 +++---- b/tools/memory-model/linux-kernel.bell | 8 ++-- b/tools/memory-model/linux-kernel.cat | 8 ++-- b/tools/memory-model/linux-kernel.def | 6 +-- b/tools/memory-model/litmus-tests/CoRR+poonceonce+Once.litmus | 7 +++ b/tools/memory-model/litmus-tests/CoRW+poonceonce+Once.litmus | 7 +++ b/tools/memory-model/litmus-tests/CoWR+poonceonce+Once.litmus | 7 +++ b/tools/memory-model/litmus-tests/CoWW+poonceonce.litmus | 7 +++ b/tools/memory-model/litmus-tests/IRIW+mbonceonces+OnceOnce.litmus | 10 +++++ b/tools/memory-model/litmus-tests/IRIW+poonceonces+OnceOnce.litmus | 10 +++++ b/tools/memory-model/litmus-tests/ISA2+poonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/ISA2+pooncerelease+poacquirerelease+poacquireonce.litmus | 11 +++++ b/tools/memory-model/litmus-tests/LB+ctrlonceonce+mbonceonce.litmus | 11 +++++ b/tools/memory-model/litmus-tests/LB+poacquireonce+pooncerelease.litmus | 8 ++++ b/tools/memory-model/litmus-tests/LB+poonceonces.litmus | 7 +++ b/tools/memory-model/litmus-tests/MP+onceassign+derefonce.litmus | 11 +++++ b/tools/memory-model/litmus-tests/MP+polocks.litmus | 11 +++++ b/tools/memory-model/litmus-tests/MP+poonceonces.litmus | 7 +++ b/tools/memory-model/litmus-tests/MP+pooncerelease+poacquireonce.litmus | 8 ++++ b/tools/memory-model/litmus-tests/MP+porevlocks.litmus | 11 +++++ b/tools/memory-model/litmus-tests/MP+wmbonceonce+rmbonceonce.litmus | 8 ++++ b/tools/memory-model/litmus-tests/R+mbonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/R+poonceonces.litmus | 8 ++++ b/tools/memory-model/litmus-tests/README | 4 +- b/tools/memory-model/litmus-tests/S+poonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/S+wmbonceonce+poacquireonce.litmus | 7 +++ b/tools/memory-model/litmus-tests/SB+mbonceonces.litmus | 9 ++++ b/tools/memory-model/litmus-tests/SB+poonceonces.litmus | 8 ++++ b/tools/memory-model/litmus-tests/WRC+poonceonces+Once.litmus | 8 ++++ b/tools/memory-model/litmus-tests/WRC+pooncerelease+rmbonceonce+Once.litmus | 8 ++++ b/tools/memory-model/litmus-tests/Z6.0+pooncelock+poonceLock+pombonce.litmus | 9 ++++ b/tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus | 8 ++++ b/tools/memory-model/litmus-tests/Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus | 14 +++++++ 36 files changed, 286 insertions(+), 38 deletions(-)
Akira Yokosawa
reviewerPATCH 1/1 tools/memory-model: Restore compat with herd7 7.47 ("-" -> "_")
Quoted 1 line
From
9a0de4d75e18ef8e25a1096f88c7a970d9d693f5Mon Sep 17 00:00:00 2001
From: Akira Yokosawa <[email protected]>
Date: Sat, 10 Feb 2018 00:45:25 +0900
Subject: [PATCH] tools/memory-model: Restore compat with herd7 7.47 ("-" -> "_")
As of herd7 7.47, these '-'s are not permitted and end up in an error:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)
Partial revert of commit 2d5fba7782d6 ("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.
--- tools/memory-model/linux-kernel.bell | 20 ++++++++++---------- tools/memory-model/linux-kernel.cat | 8 ++++---- tools/memory-model/linux-kernel.def | 14 +++++++------- 3 files changed, 21 insertions(+), 21 deletions(-)
tools/memory-model/linux-kernel.bell +10 -10
@@ -25,18 +25,18 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || 'rb_dep (*smp_read_barrier_depends*) || - 'rcu-lock (*rcu_read_lock*) || - 'rcu-unlock (*rcu_read_unlock*) || - 'sync-rcu (*synchronize_rcu*) || - 'before-atomic (*smp_mb__before_atomic*) || - 'after-atomic (*smp_mb__after_atomic*) || - 'after-spinlock (*smp_mb__after_spinlock*) + 'rcu_lock (*rcu_read_lock*) || + 'rcu_unlock (*rcu_read_unlock*) || + 'sync_rcu (*synchronize_rcu*) || + 'before_atomic (*smp_mb__before_atomic*) || + 'after_atomic (*smp_mb__after_atomic*) || + 'after_spinlock (*smp_mb__after_spinlock*) instructions F[Barriers] (* Compute matching pairs of nested Rcu-lock and Rcu-unlock *) let matched = let rec - unmatched-locks = Rcu-lock \ domain(matched) - and unmatched-unlocks = Rcu-unlock \ range(matched) + unmatched-locks = Rcu_lock \ domain(matched) + and unmatched-unlocks = Rcu_unlock \ range(matched) and unmatched = unmatched-locks | unmatched-unlocks and unmatched-po = [unmatched] ; po ; [unmatched] and unmatched-locks-to-unlocks =
@@ -46,8 +46,8 @@ let matched = let rec in matched (* Validate nesting *) -flag ~empty Rcu-lock \ domain(matched) as unbalanced-rcu-locking -flag ~empty Rcu-unlock \ range(matched) as unbalanced-rcu-locking +flag ~empty Rcu_lock \ domain(matched) as unbalanced-rcu-locking +flag ~empty Rcu_unlock \ range(matched) as unbalanced-rcu-locking (* Outermost level of nesting only *) let crit = matched \ (po^-1 ; matched ; po^-1)
tools/memory-model/linux-kernel.cat +4 -4
@@ -29,10 +29,10 @@ let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) | - ([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) | - ([M] ; po? ; [RMW] ; fencerel(After-atomic) ; [M]) | - ([M] ; po? ; [LKW] ; fencerel(After-spinlock) ; [M]) -let gp = po ; [Sync-rcu] ; po? + ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | + ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | + ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) +let gp = po ; [Sync_rcu] ; po? let strong-fence = mb | gp
tools/memory-model/linux-kernel.def +7 -7
@@ -21,9 +21,9 @@ smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } smp_read_barrier_depends() { __fence{rb_dep}; } -smp_mb__before_atomic() { __fence{before-atomic} ; } -smp_mb__after_atomic() { __fence{after-atomic} ; } -smp_mb__after_spinlock() { __fence{after-spinlock} ; } +smp_mb__before_atomic() { __fence{before_atomic} ; } +smp_mb__after_atomic() { __fence{after_atomic} ; } +smp_mb__after_spinlock() { __fence{after_spinlock} ; } // Exchange xchg(X,V) __xchg{mb}(X,V)
@@ -41,10 +41,10 @@ spin_unlock(X) { __unlock(X) ; } spin_trylock(X) __trylock(X) // RCU -rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; } // Atomic atomic_read(X) READ_ONCE(*X)
Signature · 1 line
2.7.4
Akira Yokosawa
reviewerPATCH v2 1/1 tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
In reply to Akira YokosawaQuoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001
From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)
Partial revert of commit 2d5fba7782d6 ("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.
---
Paul,
FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Thanks, Akira
--
tools/memory-model/linux-kernel.bell | 14 +++++++------- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 8 ++++---- 3 files changed, 12 insertions(+), 12 deletions(-)
tools/memory-model/linux-kernel.bell +7 -7
@@ -25,9 +25,9 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || 'rb_dep (*smp_read_barrier_depends*) || - 'rcu-lock (*rcu_read_lock*) || - 'rcu-unlock (*rcu_read_unlock*) || - 'sync-rcu (*synchronize_rcu*) || + 'rcu_lock (*rcu_read_lock*) || + 'rcu_unlock (*rcu_read_unlock*) || + 'sync_rcu (*synchronize_rcu*) || 'before_atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*)
@@ -35,8 +35,8 @@ instructions F[Barriers] (* Compute matching pairs of nested Rcu-lock and Rcu-unlock *) let matched = let rec - unmatched-locks = Rcu-lock \ domain(matched) - and unmatched-unlocks = Rcu-unlock \ range(matched) + unmatched-locks = Rcu_lock \ domain(matched) + and unmatched-unlocks = Rcu_unlock \ range(matched) and unmatched = unmatched-locks | unmatched-unlocks and unmatched-po = [unmatched] ; po ; [unmatched] and unmatched-locks-to-unlocks =
@@ -46,8 +46,8 @@ let matched = let rec in matched (* Validate nesting *) -flag ~empty Rcu-lock \ domain(matched) as unbalanced-rcu-locking -flag ~empty Rcu-unlock \ range(matched) as unbalanced-rcu-locking +flag ~empty Rcu_lock \ domain(matched) as unbalanced-rcu-locking +flag ~empty Rcu_unlock \ range(matched) as unbalanced-rcu-locking (* Outermost level of nesting only *) let crit = matched \ (po^-1 ; matched ; po^-1)
tools/memory-model/linux-kernel.cat +1 -1
@@ -32,7 +32,7 @@ let mb = ([M] ; fencerel(Mb) ; [M]) | ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) -let gp = po ; [Sync-rcu] ; po? +let gp = po ; [Sync_rcu] ; po? let strong-fence = mb | gp
tools/memory-model/linux-kernel.def +4 -4
@@ -41,10 +41,10 @@ spin_unlock(X) { __unlock(X) ; } spin_trylock(X) __trylock(X) // RCU -rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; } // Atomic atomic_read(X) READ_ONCE(*X)
Signature · 1 line
2.7.4
Re: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 16 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.
Quoted 7 lines
Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
Thanx, Paul
Quoted 82 lines
Thanks, Akira
--tools/memory-model/linux-kernel.bell | 14 +++++++------- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 8 ++++---- 3 files changed, 12 insertions(+), 12 deletions(-)tools/memory-model/linux-kernel.bell +7
-7@@ -25,9 +25,9 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || 'rb_dep (*smp_read_barrier_depends*) || - 'rcu-lock (*rcu_read_lock*) || - 'rcu-unlock (*rcu_read_unlock*) || - 'sync-rcu (*synchronize_rcu*) || + 'rcu_lock (*rcu_read_lock*) || + 'rcu_unlock (*rcu_read_unlock*) || + 'sync_rcu (*synchronize_rcu*) || 'before_atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*)@@ -35,8 +35,8 @@ instructions F[Barriers] (* Compute matching pairs of nested Rcu-lock and Rcu-unlock *) let matched = let rec - unmatched-locks = Rcu-lock \ domain(matched) - and unmatched-unlocks = Rcu-unlock \ range(matched) + unmatched-locks = Rcu_lock \ domain(matched) + and unmatched-unlocks = Rcu_unlock \ range(matched) and unmatched = unmatched-locks | unmatched-unlocks and unmatched-po = [unmatched] ; po ; [unmatched] and unmatched-locks-to-unlocks =@@ -46,8 +46,8 @@ let matched = let rec in matched (* Validate nesting *) -flag ~empty Rcu-lock \ domain(matched) as unbalanced-rcu-locking -flag ~empty Rcu-unlock \ range(matched) as unbalanced-rcu-locking +flag ~empty Rcu_lock \ domain(matched) as unbalanced-rcu-locking +flag ~empty Rcu_unlock \ range(matched) as unbalanced-rcu-locking (* Outermost level of nesting only *) let crit = matched \ (po^-1 ; matched ; po^-1)tools/memory-model/linux-kernel.cat +1
-1@@ -32,7 +32,7 @@ let mb = ([M] ; fencerel(Mb) ; [M]) | ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) -let gp = po ; [Sync-rcu] ; po? +let gp = po ; [Sync_rcu] ; po? let strong-fence = mb | gptools/memory-model/linux-kernel.def +4
-4@@ -41,10 +41,10 @@ spin_unlock(X) { __unlock(X) ; } spin_trylock(X) __trylock(X) // RCU -rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; } // Atomic atomic_read(X) READ_ONCE(*X)Signature · 1 line
2.7.4
Akira Yokosawa
reviewerRe: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 23 lines · Paul E. McKenney
Quoted 22 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,
Quoted 7 lines
Quoted 1 line
FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
Fair enough.
Thanks, Akira
Quoted 86 lines
Thanx, Paul
Quoted 82 lines
Thanks, Akira
--tools/memory-model/linux-kernel.bell | 14 +++++++------- tools/memory-model/linux-kernel.cat | 2 +- tools/memory-model/linux-kernel.def | 8 ++++---- 3 files changed, 12 insertions(+), 12 deletions(-)tools/memory-model/linux-kernel.bell +7
-7@@ -25,9 +25,9 @@ enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || 'rb_dep (*smp_read_barrier_depends*) || - 'rcu-lock (*rcu_read_lock*) || - 'rcu-unlock (*rcu_read_unlock*) || - 'sync-rcu (*synchronize_rcu*) || + 'rcu_lock (*rcu_read_lock*) || + 'rcu_unlock (*rcu_read_unlock*) || + 'sync_rcu (*synchronize_rcu*) || 'before_atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*)@@ -35,8 +35,8 @@ instructions F[Barriers] (* Compute matching pairs of nested Rcu-lock and Rcu-unlock *) let matched = let rec - unmatched-locks = Rcu-lock \ domain(matched) - and unmatched-unlocks = Rcu-unlock \ range(matched) + unmatched-locks = Rcu_lock \ domain(matched) + and unmatched-unlocks = Rcu_unlock \ range(matched) and unmatched = unmatched-locks | unmatched-unlocks and unmatched-po = [unmatched] ; po ; [unmatched] and unmatched-locks-to-unlocks =@@ -46,8 +46,8 @@ let matched = let rec in matched (* Validate nesting *) -flag ~empty Rcu-lock \ domain(matched) as unbalanced-rcu-locking -flag ~empty Rcu-unlock \ range(matched) as unbalanced-rcu-locking +flag ~empty Rcu_lock \ domain(matched) as unbalanced-rcu-locking +flag ~empty Rcu_unlock \ range(matched) as unbalanced-rcu-locking (* Outermost level of nesting only *) let crit = matched \ (po^-1 ; matched ; po^-1)tools/memory-model/linux-kernel.cat +1
-1@@ -32,7 +32,7 @@ let mb = ([M] ; fencerel(Mb) ; [M]) | ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) -let gp = po ; [Sync-rcu] ; po? +let gp = po ; [Sync_rcu] ; po? let strong-fence = mb | gptools/memory-model/linux-kernel.def +4
-4@@ -41,10 +41,10 @@ spin_unlock(X) { __unlock(X) ; } spin_trylock(X) __trylock(X) // RCU -rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; } // Atomic atomic_read(X) READ_ONCE(*X)Signature · 1 line
2.7.4
Re: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
In reply to Paul E. McKenneyQuoted 23 lines · Paul E. McKenney
Quoted 22 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,
Quoted 7 lines
Quoted 1 line
FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
I'm wondering:
Quoted 9 lines
tools/memory-model/linux-kernel.def spin_unlock(X) { __unlock(X) ; } · line 47
Quoted 9 lines
// RCU
tools/memory-model/linux-kernel.def +4
-4-rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; }
What's the point of using '-' instead of '_'? In a program language syntax environment it's easy to confuse it with a '-' operator, and it also looks slightly sloppy and inconsistent if we sometimes have '_' and sometimes '-'.
If it used on purpose, to separate namespaces from kernel internal API names, then that's rather obscure IMHO and then I'd rather suggest proper prefixes instead.
Thanks,
Ingo
Re: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 50 lines · Ingo Molnar
Quoted 23 lines · Paul E. McKenney
Quoted 22 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,Quoted 7 lines
Quoted 1 line
FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
I'm wondering:
Quoted 9 lines
tools/memory-model/linux-kernel.def spin_unlock(X) { __unlock(X) ; } · line 47
Quoted 9 lines
// RCU
tools/memory-model/linux-kernel.def +4
-4-rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; }What's the point of using '-' instead of '_'? In a program language syntax environment it's easy to confuse it with a '-' operator, and it also looks slightly sloppy and inconsistent if we sometimes have '_' and sometimes '-'.
Quoted 3 lines
If it used on purpose, to separate namespaces from kernel internal API names, then that's rather obscure IMHO and then I'd rather suggest proper prefixes instead.
These lines are connecting Linux-kernel names to cat-language names. The cat language can be thought of as a set-based constraint language, which doesn't have subtraction (the closest analog being set difference, which is denoted by "\"). Cat-language code has used "-" as a word separator in names, and the model had previously had an odd mixture of "-" and "_". Hence, the switch to the cat-language convention.
Of course, if underscore is strongly preferred, underscore does work, and it should be possible to move in that direction.
Thanx, Paul
Re: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 6 lines · Paul E. McKenney
Cat-language code has used "-" as a word
separator in names, and the model had previously had an odd mixture of "-" and "_". Hence, the switch to the cat-language convention.Of course, if underscore is strongly preferred, underscore does work, and it should be possible to move in that direction.
I'm fine with either of these variants, as long as it's consistent: the idiosynchratic seeming mixture of both '-' and '_' convention looked odd to me.
If '_' is also more compatible with installed versions of the language, then I'd go for that - but it's your call.
Thanks,
Ingo
Akira Yokosawa
reviewerRe: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
In reply to Paul E. McKenneyQuoted 23 lines · Paul E. McKenney
Quoted 22 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,
Quoted 7 lines
Quoted 1 line
FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
So, herdtools7.7.48 is now available by "opam update; opam upgrade". Maybe mentioning the required version of herdtools7 in README would help.
One glitch I'm aware of is one of the output of klitmus7 fails to compile on kernels prior to 4.14, because of smp_mb__after_spinlock() used in Z6.0+pooncelock+poonceLock+pombonce.litmus.
And there is a question I want to ask.
Z6.0+pooncelock+pooncelock+pombonce.litmus, which doesn't have smp_mb__after_spinlock(), has a comment of:
(*
* Result: Never
*
* This example demonstrates that a pair of accesses made by different
* processes each while holding a given lock will not necessarily be
* seen as ordered by a third process not holding that lock. *)
The result was changed from "Sometimes" to "Never" in response to Alan's feedback the other day, but the following comment indicates the result should be "Sometimes". And the current model gives the result of "Sometimes".
As I have no idea of the pending patches, I'd like to know once the pending patches are applied, would Z6.0+pooncelock+pooncelock+pombonce.litmus and Z6.0+pooncelock+poonceLock+pombonce.litmus become identical?
Thanks, Akira
Quoted 2 lines
Thanx, Paul
[...]
d_lock*) ||
Quoted 64 lines
Quoted 63 lines
tools/memory-model/linux-kernel.bell +6
-4+ 'rcu_unlock (*rcu_read_unlock*) || + 'sync_rcu (*synchronize_rcu*) || 'before_atomic (*smp_mb__before_atomic*) || 'after_atomic (*smp_mb__after_atomic*) || 'after_spinlock (*smp_mb__after_spinlock*)@@ -35,8 +35,8 @@ instructions F[Barriers] (* Compute matching pairs of nested Rcu-lock and Rcu-unlock *) let matched = let rec - unmatched-locks = Rcu-lock \ domain(matched) - and unmatched-unlocks = Rcu-unlock \ range(matched) + unmatched-locks = Rcu_lock \ domain(matched) + and unmatched-unlocks = Rcu_unlock \ range(matched) and unmatched = unmatched-locks | unmatched-unlocks and unmatched-po = [unmatched] ; po ; [unmatched] and unmatched-locks-to-unlocks =@@ -46,8 +46,8 @@ let matched = let rec in matched (* Validate nesting *) -flag ~empty Rcu-lock \ domain(matched) as unbalanced-rcu-locking -flag ~empty Rcu-unlock \ range(matched) as unbalanced-rcu-locking +flag ~empty Rcu_lock \ domain(matched) as unbalanced-rcu-locking +flag ~empty Rcu_unlock \ range(matched) as unbalanced-rcu-locking (* Outermost level of nesting only *) let crit = matched \ (po^-1 ; matched ; po^-1)tools/memory-model/linux-kernel.cat +1
-1@@ -32,7 +32,7 @@ let mb = ([M] ; fencerel(Mb) ; [M]) | ([M] ; fencerel(Before_atomic) ; [RMW] ; po? ; [M]) | ([M] ; po? ; [RMW] ; fencerel(After_atomic) ; [M]) | ([M] ; po? ; [LKW] ; fencerel(After_spinlock) ; [M]) -let gp = po ; [Sync-rcu] ; po? +let gp = po ; [Sync_rcu] ; po? let strong-fence = mb | gptools/memory-model/linux-kernel.def +4
-4@@ -41,10 +41,10 @@ spin_unlock(X) { __unlock(X) ; } spin_trylock(X) __trylock(X) // RCU -rcu_read_lock() { __fence{rcu-lock}; } -rcu_read_unlock() { __fence{rcu-unlock};} -synchronize_rcu() { __fence{sync-rcu}; } -synchronize_rcu_expedited() { __fence{sync-rcu}; } +rcu_read_lock() { __fence{rcu_lock}; } +rcu_read_unlock() { __fence{rcu_unlock};} +synchronize_rcu() { __fence{sync_rcu}; } +synchronize_rcu_expedited() { __fence{sync_rcu}; } // Atomic atomic_read(X) READ_ONCE(*X)Signature · 1 line
2.7.4
Re: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 26 lines · Akira Yokosawa
Quoted 25 lines · Paul E. McKenney
Quoted 23 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Quoted 8 lines
Quoted 5 lines
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
So, herdtools7.7.48 is now available by "opam update; opam upgrade". Maybe mentioning the required version of herdtools7 in README would help.
Or have some way for the memory model's .cat files to state what version they need, but in the meantime please see the patch below. But even with such version specification, we probably want to have the version in the README...
Quoted 3 lines
One glitch I'm aware of is one of the output of klitmus7 fails to compile on kernels prior to 4.14, because of
smp_mb__after_spinlock()used in Z6.0+pooncelock+poonceLock+pombonce.litmus.
This is one advantage of having the memory model in the kernel source itself -- the versions match. And people can always fire up a different kernel version (for example, within a VM) to run the output of klitmus7.
Quoted 18 lines
And there is a question I want to ask.
Z6.0+pooncelock+pooncelock+pombonce.litmus, which doesn't have
smp_mb__after_spinlock(), has a comment of:(*
* Result: Never
*
* This example demonstrates that a pair of accesses made by different
* processes each while holding a given lock will not necessarily be
* seen as ordered by a third process not holding that lock. *)The result was changed from "Sometimes" to "Never" in response to Alan's feedback the other day, but the following comment indicates the result should be "Sometimes". And the current model gives the result of "Sometimes".
Quoted 4 lines
As I have no idea of the pending patches, I'd like to know once the pending patches are applied, would Z6.0+pooncelock+pooncelock+pombonce.litmus and Z6.0+pooncelock+poonceLock+pombonce.litmus become identical?
Let's see, examining the Z6 litmus tests and running on herd7 version 7.48:
Z6.0+pooncelock+pooncelock+pombonce Sometimes 1 7 Z6.0+pooncelock+poonceLock+pombonce Never 0 7 Z6.0+pooncerelease+poacquirerelease+mbonceonce Sometimes 1 7
But Alan stated that the result of the test would be changed by one of the patches in our "pending" list. I ran with all the patches currently applied, so I am guessing that Alan was refering to one of the changes that we have discussed but not yet created.
Alan, enlightenment?
Thanx, Paul
------------------------------------------------------------------------
commit dad94d8dc78e50bab91b313a67232e85ee22a733
Author: Paul E. McKenney <[email protected]> Date: Wed Feb 14 14:37:22 2018 -0800
tools/memory-model: Add required herd7 version to README file
LKMM and the herd7 tool are co-evolving, and out-of-date herd7 tools produce inaccurate results, often with no obvious error messages. This commit therefore adds the required herd7 version to the LKMM README file.
Longer term, it would be good if .cat files could specify the required version in a manner allowing herd7 to produce clear diagnostics.
tools/memory-model/README +2 -1
@@ -20,7 +20,8 @@ that litmus test to be exercised within the Linux kernel. REQUIREMENTS ============ -The "herd7" and "klitmus7" tools must be downloaded separately: +Version 7.48 of the "herd7" and "klitmus7" tools must be downloaded +separately: https://github.com/herd/herdtools7
Alan Stern
maintainerRe: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 6 lines · Paul E. McKenney
Let's see, examining the Z6 litmus tests and running on herd7 version 7.48:
Z6.0+pooncelock+pooncelock+pombonce Sometimes 1 7 Z6.0+pooncelock+poonceLock+pombonce Never 0 7 Z6.0+pooncerelease+poacquirerelease+mbonceonce Sometimes 1 7
Quoted 6 lines
But Alan stated that the result of the test would be changed by one of the patches in our "pending" list. I ran with all the patches currently applied, so I am guessing that Alan was refering to one of the changes that we have discussed but not yet created.
Alan, enlightenment?
Here's the relevant patch. It may need some manual adjustment, because it was not made against the files currently in the repository.
Alan
linux-kernel-hardware.cat +0 -0
linux-kernel-hardware.cat +4 -4
@@ -31,7 +31,7 @@ let strong-fence = mb | gp (* Release Acquire *) let acq-po = [Acquire] ; po ; [M] let po-rel = [M] ; po ; [Release] -let rfi-rel-acq = [Release] ; rfi ; [Acquire] +let rel-rf-acq-po = [Release] ; rf ; [Acquire] ; po (**********************************) (* Fundamental coherence ordering *)
@@ -57,13 +57,13 @@ let to-w = rwdep | addrpo | (overwrite & let rdw = po-loc & (fre ; rfe) let detour = po-loc & (coe ; rfe) let rrdep = addr | (dep ; rfi) -let to-r = rrdep | rfi-rel-acq -let fence = strong-fence | wmb | po-rel | rmb | acq-po +let to-r = rrdep +let fence = strong-fence | wmb | po-rel | rmb | acq-po | (rel-rf-acq-po & int) let ppo = to-r | to-w | fence | rdw | detour (* Happens Before *) let A-cumul(r) = rfe? ; r -let cumul-fence = A-cumul(strong-fence | po-rel) | wmb +let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po let rec prop = (overwrite & ext)? ; cumul-fence ; hb* and hb = ppo | rfe | ((hb* ; prop) & int)
linux-kernel.cat +4 -4
@@ -31,7 +31,7 @@ let strong-fence = mb | gp (* Release Acquire *) let acq-po = [Acquire] ; po ; [M] let po-rel = [M] ; po ; [Release] -let rfi-rel-acq = [Release] ; rfi ; [Acquire] +let rel-rf-acq-po = [Release] ; rf ; [Acquire] ; po (**********************************) (* Fundamental coherence ordering *)
@@ -54,13 +54,13 @@ let rwdep = (dep | ctrl) ; [W] let overwrite = co | fr let to-w = rwdep | (overwrite & int) let rrdep = addr | (dep ; rfi) -let to-r = rrdep | rfi-rel-acq -let fence = strong-fence | wmb | po-rel | rmb | acq-po +let to-r = rrdep +let fence = strong-fence | wmb | po-rel | rmb | acq-po | (rel-rf-acq-po & int) let ppo = to-r | to-w | fence (* Propagation: Ordering from release operations and strong fences. *) let A-cumul(r) = rfe? ; r -let cumul-fence = A-cumul(strong-fence | po-rel) | wmb +let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po let prop = (overwrite & ext)? ; cumul-fence* ; rfe? (*
Akira Yokosawa
reviewerTrial of conflict resolution of Alan's patch
Quoted 1 line
From
81bd6374057f0d89894ead482b870e2f001d2998Mon Sep 17 00:00:00 2001
From: Alan Stern <[email protected]>
Date: Fri, 16 Feb 2018 00:29:56 +0900
Subject: [PATCH] [TENTATIVE] Trial conflict resolution of Alan's patch
This is a trial of conflict resolution of the patch Alan provided.
Alan's message and original patch:
Here's the relevant patch. It may need some manual adjustment, because it was not made against the files currently in the repository.
Alan
diff --git a/linux-kernel-hardware.cat b/linux-kernel-hardware.cat
Index: memory-model/linux-kernel-hardware.cat
===================================================================
--- memory-model.orig/linux-kernel-hardware.cat
+++ memory-model/linux-kernel-hardware.cat
@@ -31,7 +31,7 @@ let strong-fence = mb | gp
(* Release Acquire *)
let acq-po = [Acquire] ; po ; [M]
let po-rel = [M] ; po ; [Release]
-let rfi-rel-acq = [Release] ; rfi ; [Acquire]
+let rel-rf-acq-po = [Release] ; rf ; [Acquire] ; po
(**********************************)
(* Fundamental coherence ordering *)
@@ -57,13 +57,13 @@ let to-w = rwdep | addrpo | (overwrite &
let rdw = po-loc & (fre ; rfe)
let detour = po-loc & (coe ; rfe)
let rrdep = addr | (dep ; rfi)
-let to-r = rrdep | rfi-rel-acq
-let fence = strong-fence | wmb | po-rel | rmb | acq-po
+let to-r = rrdep
+let fence = strong-fence | wmb | po-rel | rmb | acq-po | (rel-rf-acq-po & int)
let ppo = to-r | to-w | fence | rdw | detour
(* Happens Before *)
let A-cumul(r) = rfe? ; r
-let cumul-fence = A-cumul(strong-fence | po-rel) | wmb +let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po
let rec prop = (overwrite & ext)? ; cumul-fence ; hb*
and hb = ppo | rfe | ((hb* ; prop) & int)
Index: memory-model/linux-kernel.cat
===================================================================
--- memory-model.orig/linux-kernel.cat
+++ memory-model/linux-kernel.cat
@@ -31,7 +31,7 @@ let strong-fence = mb | gp
(* Release Acquire *)
let acq-po = [Acquire] ; po ; [M]
let po-rel = [M] ; po ; [Release]
-let rfi-rel-acq = [Release] ; rfi ; [Acquire] +let rel-rf-acq-po = [Release] ; rf ; [Acquire] ; po
(**********************************)
(* Fundamental coherence ordering *)
@@ -54,13 +54,13 @@ let rwdep = (dep | ctrl) ; [W]
let overwrite = co | fr
let to-w = rwdep | (overwrite & int)
let rrdep = addr | (dep ; rfi)
-let to-r = rrdep | rfi-rel-acq
-let fence = strong-fence | wmb | po-rel | rmb | acq-po
+let to-r = rrdep
+let fence = strong-fence | wmb | po-rel | rmb | acq-po | (rel-rf-acq-po & int)
let ppo = to-r | to-w | fence
(* Propagation: Ordering from release operations and strong fences. *) let A-cumul(r) = rfe? ; r
-let cumul-fence = A-cumul(strong-fence | po-rel) | wmb +let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po
let prop = (overwrite & ext)? ; cumul-fence* ; rfe?
(*
---
So, I attempted to rebase the patch to current (somewhat old) master of
https://github.com/aparri/memory-model. Why? Because the lkmm branch in Paul's -rcu tree doesn't have linux-kernel-hardware.cat.
However, after this change, Z6.0+pooncelock+pooncelock+pombonce still has the result "Sometimes". I must have done something wrong in the conflict resolution.
Note: I have almost no idea what this patch is doing. I'm just hoping to give a starting point of a discussion.
Thanks, Akira
--
linux-kernel-hardware.cat | 9 ++++----- linux-kernel.cat | 9 ++++----- 2 files changed, 8 insertions(+), 10 deletions(-)
linux-kernel-hardware.cat +4 -5
@@ -34,7 +34,7 @@ let strong-fence = mb | gp (* Release Acquire *) let acq-po = [Acquire] ; po ; [M] let po-rel = [M] ; po ; [Release] -let rfi-rel-acq = [Release] ; rfi ; [Acquire] +let rel-rf-acq-po = [Release] ; rf ; [Acquire] ; po (**********************************) (* Fundamental coherence ordering *)
@@ -60,14 +60,13 @@ let to-w = rwdep | addrpo | (overwrite & int) let rdw = po-loc & (fre ; rfe) let detour = po-loc & (coe ; rfe) let rrdep = addr | (dep ; rfi) | rdw -let strong-rrdep = rrdep+ & rb-dep -let to-r = strong-rrdep | rfi-rel-acq -let fence = strong-fence | wmb | po-rel | rmb | acq-po +let to-r = rrdep +let fence = strong-fence | wmb | po-rel | rmb | acq-po | (rel-rf-acq-po & int) let ppo = (rrdep* ; (to-r | to-w | fence)) | rdw | detour (* Happens Before *) let A-cumul(r) = rfe? ; r -let cumul-fence = A-cumul(strong-fence | po-rel) | wmb +let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po let rec prop = (overwrite & ext)? ; cumul-fence ; hb* and hb = ppo | rfe | ((hb* ; prop) & int)
linux-kernel.cat +4 -5
@@ -39,7 +39,7 @@ let strong-fence = mb | gp (* Release Acquire *) let acq-po = [Acquire] ; po ; [M] let po-rel = [M] ; po ; [Release] -let rfi-rel-acq = [Release] ; rfi ; [Acquire] +let rel-rf-acq-po = [Release] ; rf ; [Acquire] ; po (**********************************) (* Fundamental coherence ordering *)
@@ -62,14 +62,13 @@ let rwdep = (dep | ctrl) ; [W] let overwrite = co | fr let to-w = rwdep | (overwrite & int) let rrdep = addr | (dep ; rfi) -let strong-rrdep = rrdep+ & rb-dep -let to-r = strong-rrdep | rfi-rel-acq -let fence = strong-fence | wmb | po-rel | rmb | acq-po +let to-r = rrdep +let fence = strong-fence | wmb | po-rel | rmb | acq-po | (rel-rf-acq-po & int) let ppo = rrdep* ; (to-r | to-w | fence) (* Propagation: Ordering from release operations and strong fences. *) let A-cumul(r) = rfe? ; r -let cumul-fence = A-cumul(strong-fence | po-rel) | wmb +let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | rel-rf-acq-po let prop = (overwrite & ext)? ; cumul-fence* ; rfe? (*
Signature · 1 line
2.7.4
Alan Stern
maintainerRe: Trial of conflict resolution of Alan's patch
Quoted 10 lines · Akira Yokosawa
So, I attempted to rebase the patch to current (somewhat old) master of
https://github.com/aparri/memory-model. Why? Because the lkmm branch in Paul's -rcu tree doesn't have linux-kernel-hardware.cat.However, after this change, Z6.0+pooncelock+pooncelock+pombonce still has the result "Sometimes". I must have done something wrong in the conflict resolution.
Note: I have almost no idea what this patch is doing. I'm just hoping to give a starting point of a discussion.
Yes, that litmus test gives "Sometimes" both with and without the patch. But consider instead this slightly changed version of that test, in which P2 reads Z instead of writing it:
C Z6.0-variant
{}
P0(int *x, int *y, spinlock_t *mylock)
{
spin_lock(mylock);
WRITE_ONCE(*x, 1);
WRITE_ONCE(*y, 1);
spin_unlock(mylock);
}P1(int *y, int *z, spinlock_t *mylock)
{
int r0;
spin_lock(mylock);
r0 = READ_ONCE(*y);
WRITE_ONCE(*z, 1);
spin_unlock(mylock);
}P2(int *x, int *z)
{
int r1;
int r2;
r2 = READ_ONCE(*z);
smp_mb();
r1 = READ_ONCE(*x);
}exists (1:r0=1 /\ 2:r2=1 /\ 2:r1=0)
Without the patch, this test gives "Sometimes"; with the patch it gives "Never". That is what I thought Paul was talking about originally.
Sorry if my misunderstanding caused too much confusion for other people.
Alan
Re: Trial of conflict resolution of Alan's patch
Quoted 49 lines · Alan Stern
Quoted 10 lines · Akira Yokosawa
So, I attempted to rebase the patch to current (somewhat old) master of
https://github.com/aparri/memory-model. Why? Because the lkmm branch in Paul's -rcu tree doesn't have linux-kernel-hardware.cat.However, after this change, Z6.0+pooncelock+pooncelock+pombonce still has the result "Sometimes". I must have done something wrong in the conflict resolution.
Note: I have almost no idea what this patch is doing. I'm just hoping to give a starting point of a discussion.
Yes, that litmus test gives "Sometimes" both with and without the patch. But consider instead this slightly changed version of that test, in which P2 reads Z instead of writing it:
C Z6.0-variant
{}
P0(int *x, int *y, spinlock_t *mylock){ spin_lock(mylock); WRITE_ONCE(*x, 1); WRITE_ONCE(*y, 1); spin_unlock(mylock); }
P1(int *y, int *z, spinlock_t *mylock){ int r0; spin_lock(mylock); r0 = READ_ONCE(*y); WRITE_ONCE(*z, 1); spin_unlock(mylock); }
P2(int *x, int *z){ int r1; int r2; r2 = READ_ONCE(*z); smp_mb(); r1 = READ_ONCE(*x); }
Quoted 7 lines
exists (1:r0=1 /\ 2:r2=1 /\ 2:r1=0)
Without the patch, this test gives "Sometimes"; with the patch it gives "Never". That is what I thought Paul was talking about originally.
Sorry if my misunderstanding caused too much confusion for other people.
Ah, I did indeed get confused. I have changed the "Result:" for Z6.0+pooncelock+pooncelock+pombonce.litmus back to "Never", as in the patch below (which I merged into the patch adding all the comments).
I have added the above test as ISA2+pooncelock+pooncelock+pombonce.litmus, with the Result: of Sometimes with you (Alan) as author and with your Signed-off-by -- please let me know if you would prefer some other approach.
Please change the Result: when sending the proposed patch. Or please let me know if you would like me to apply the forward-port that Akira sent, in which case I will add the Result: change to that patch. Or for that matter, Akira might repost his forward-port of your patch with this change.
Thanx, Paul
------------------------------------------------------------------------
commit b2950420e1154131c0667f1ac58666bad3a06a69
Author: Paul E. McKenney <[email protected]> Date: Thu Feb 15 10:35:25 2018 -0800
fixup! EXP litmus_tests: Add comments explaining tests' purposes
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +1 -1
@@ -1,7 +1,7 @@ C Z6.0+pooncelock+pooncelock+pombonce (* - * Result: Never + * Result: Somtimes * * This example demonstrates that a pair of accesses made by different * processes each while holding a given lock will not necessarily be
Akira Yokosawa
reviewerRe: Trial of conflict resolution of Alan's patch
Quoted 68 lines · Paul E. McKenney
Quoted 49 lines · Alan Stern
Quoted 10 lines · Akira Yokosawa
So, I attempted to rebase the patch to current (somewhat old) master of
https://github.com/aparri/memory-model. Why? Because the lkmm branch in Paul's -rcu tree doesn't have linux-kernel-hardware.cat.However, after this change, Z6.0+pooncelock+pooncelock+pombonce still has the result "Sometimes". I must have done something wrong in the conflict resolution.
Note: I have almost no idea what this patch is doing. I'm just hoping to give a starting point of a discussion.
Yes, that litmus test gives "Sometimes" both with and without the patch. But consider instead this slightly changed version of that test, in which P2 reads Z instead of writing it:
C Z6.0-variant
{}
P0(int *x, int *y, spinlock_t *mylock){ spin_lock(mylock); WRITE_ONCE(*x, 1); WRITE_ONCE(*y, 1); spin_unlock(mylock); }
P1(int *y, int *z, spinlock_t *mylock){ int r0; spin_lock(mylock); r0 = READ_ONCE(*y); WRITE_ONCE(*z, 1); spin_unlock(mylock); }
P2(int *x, int *z){ int r1; int r2; r2 = READ_ONCE(*z); smp_mb(); r1 = READ_ONCE(*x); }Quoted 7 lines
exists (1:r0=1 /\ 2:r2=1 /\ 2:r1=0)
Without the patch, this test gives "Sometimes"; with the patch it gives "Never". That is what I thought Paul was talking about originally.
Sorry if my misunderstanding caused too much confusion for other people.
Ah, I did indeed get confused. I have changed the "Result:" for Z6.0+pooncelock+pooncelock+pombonce.litmus back to "Never", as in the patch below (which I merged into the patch adding all the comments).
I have added the above test as ISA2+pooncelock+pooncelock+pombonce.litmus, with the Result: of Sometimes with you (Alan) as author and with your Signed-off-by -- please let me know if you would prefer some other approach.
Quoted 5 lines
Please change the Result: when sending the proposed patch. Or please let me know if you would like me to apply the forward-port that Akira sent, in which case I will add the Result: change to that patch. Or for that matter, Akira might repost his forward-port of your patch with this change.
Hi,
My forward-port patch doesn't apply to the "lkmm" branch.
It looks like "linux-kernel-hardware.cat" is intentionally omitted there. Am I guessing right?
If this is the case, I can prepare a patch to be applied to "lkmm". But I can't compose a proper change log. So I'd like Alan to post a patch with my SOB appended. Does this approach sound reasonable?
Thanks, Akira
Quoted 26 lines
Thanx, Paul
------------------------------------------------------------------------
commit
b2950420e1154131c0667f1ac58666bad3a06a69Author: Paul E. McKenney <[email protected]> Date: Thu Feb 15 10:35:25 2018 -0800fixup! EXP
litmus_tests: Add comments explaining tests' purposesSigned-off-by Paul E. McKenney <[email protected]>tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +1
-1@@ -1,7 +1,7 @@ C Z6.0+pooncelock+pooncelock+pombonce (* - * Result: Never + * Result: Somtimes * * This example demonstrates that a pair of accesses made by different * processes each while holding a given lock will not necessarily be
Alan Stern
maintainerRe: Trial of conflict resolution of Alan's patch
Quoted 9 lines · Akira Yokosawa
Hi,
My forward-port patch doesn't apply to the "lkmm" branch.
It looks like "linux-kernel-hardware.cat" is intentionally omitted there. Am I guessing right?If this is the case, I can prepare a patch to be applied to "lkmm". But I can't compose a proper change log. So I'd like Alan to post a patch with my SOB appended. Does this approach sound reasonable?
The patch is not yet ready to be merged. At the very least, I need to include an update to explanation.txt along with it. When it is all ready, I will rebase it on Paul's repository and post it.
Which reminds me: Now that the material has been accepted into the kernel, do we need to keep the github repository? It has the linux-kernel-hardware.cat file, but otherwise it seems to be redundant.
Alan
Andrea Parri
maintainerRe: Trial of conflict resolution of Alan's patch
Quoted 12 lines · Alan Stern
Quoted 9 lines · Akira Yokosawa
Hi,
My forward-port patch doesn't apply to the "lkmm" branch.
It looks like "linux-kernel-hardware.cat" is intentionally omitted there. Am I guessing right?If this is the case, I can prepare a patch to be applied to "lkmm". But I can't compose a proper change log. So I'd like Alan to post a patch with my SOB appended. Does this approach sound reasonable?
Quoted 7 lines
The patch is not yet ready to be merged. At the very least, I need to include an update to explanation.txt along with it. When it is all ready, I will rebase it on Paul's repository and post it.
Which reminds me: Now that the material has been accepted into the kernel, do we need to keep the github repository? It has the linux-kernel-hardware.cat file, but otherwise it seems to be redundant.
If you mean "to keep up-to-date", I'd say "No, we don't..." ;-)
My plan/hope is to add such a disclaimer together with a pointer to Linus's tree ASAP...
Andrea
Quoted 3 lines
Alan
Re: Trial of conflict resolution of Alan's patch
Quoted 17 lines · Andrea Parri
Quoted 16 lines · Alan Stern
Quoted 9 lines · Akira Yokosawa
Hi,
My forward-port patch doesn't apply to the "lkmm" branch.
It looks like "linux-kernel-hardware.cat" is intentionally omitted there. Am I guessing right?If this is the case, I can prepare a patch to be applied to "lkmm". But I can't compose a proper change log. So I'd like Alan to post a patch with my SOB appended. Does this approach sound reasonable?
The patch is not yet ready to be merged. At the very least, I need to include an update to explanation.txt along with it. When it is all ready, I will rebase it on Paul's repository and post it.
Quoted 8 lines
Quoted 3 lines
Which reminds me: Now that the material has been accepted into the kernel, do we need to keep the github repository? It has the linux-kernel-hardware.cat file, but otherwise it seems to be redundant.
If you mean "to keep up-to-date", I'd say "No, we don't..." ;-)
My plan/hope is to add such a disclaimer together with a pointer to Linus's tree ASAP...
I agree that the github repository is useful as a historical reference but that we should not try to keep it up to date. The -rc1 release came out last Sunday, so in 7-8 weeks we will hopefully have this in the Linux kernel. Fingers firmly crossed and all that...
Thanx, Paul
Alan Stern
maintainerPATCH 1/1 tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Since commit 76ebbe78f739 ("locking/barriers: Add implicit smp_read_barrier_depends() to READ_ONCE()") was merged for the 4.15 kernel, it has not been necessary to use smp_read_barrier_depends(). Similarly, commit 59ecbbe7b31c ("locking/barriers: Kill lockless_dereference()") removed lockless_dereference() from the kernel.
Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
---
tools/memory-model/linux-kernel.cat renamed from tools/memory-model.orig/linux-kernel.cat +2 -5
@@ -25,7 +25,6 @@ include "lock.cat" (*******************) (* Fences *) -let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) |
@@ -61,11 +60,9 @@ let dep = addr | data let rwdep = (dep | ctrl) ; [W] let overwrite = co | fr let to-w = rwdep | (overwrite & int) -let rrdep = addr | (dep ; rfi) -let strong-rrdep = rrdep+ & rb-dep -let to-r = strong-rrdep | rfi-rel-acq +let to-r = addr | (dep ; rfi) | rfi-rel-acq let fence = strong-fence | wmb | po-rel | rmb | acq-po -let ppo = rrdep* ; (to-r | to-w | fence) +let ppo = to-r | to-w | fence (* Propagation: Ordering from release operations and strong fences. *) let A-cumul(r) = rfe? ; r
tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +43 -38
144 lines not on this page · raw message
tools/memory-model/linux-kernel.bell renamed from tools/memory-model.orig/linux-kernel.bell +0 -1
@@ -24,7 +24,6 @@ instructions RMW[{'once,'acquire,'releas enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || - 'rb_dep (*smp_read_barrier_depends*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) ||
tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0 -2
@@ -13,14 +13,12 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -lockless_dereference(X) __load{lderef}(X) rcu_dereference(X) __load{deref}(X) // Fences smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } -smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after-atomic} ; } smp_mb__after_spinlock() { __fence{after-spinlock} ; }
tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +1 -2
@@ -6,8 +6,7 @@ Store, e.g., WRITE_ONCE() Y Y Load, e.g., READ_ONCE() Y Y Y Unsuccessful RMW operation Y Y Y -smp_read_barrier_depends() Y Y Y -*_dereference() Y Y Y Y +rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Y smp_rmb() Y R Y Y R
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 7 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.
Quoted 7 lines
Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Signed-off-by Alan Stern <[email protected]>
A few nits. Please see inline comments below.
With those fixed,
Thanks, Akira
Quoted 81 lines
---
Index: usb-4.x/tools/memory-model/linux-kernel.cat
===================================================================tools/memory-model/linux-kernel.cat renamed from tools/memory-model.orig/linux-kernel.cat +2
-5@@ -25,7 +25,6 @@ include "lock.cat" (*******************) (* Fences *) -let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) |@@ -61,11 +60,9 @@ let dep = addr | data let rwdep = (dep | ctrl) ; [W] let overwrite = co | fr let to-w = rwdep | (overwrite & int) -let rrdep = addr | (dep ; rfi) -let strong-rrdep = rrdep+ & rb-dep -let to-r = strong-rrdep | rfi-rel-acq +let to-r = addr | (dep ; rfi) | rfi-rel-acq let fence = strong-fence | wmb | po-rel | rmb | acq-po -let ppo = rrdep* ; (to-r | to-w | fence) +let ppo = to-r | to-w | fence (* Propagation: Ordering from release operations and strong fences. *) let A-cumul(r) = rfe? ; rIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +16
-17@@ -1,5 +1,5 @@ -Explanation of the Linux-Kernel Memory Model -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Explanation of the Linux-Kernel Memory Consistency Model +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ :Author: Alan Stern <[email protected]> :Created: October 2017@@ -35,25 +35,24 @@ Explanation of the Linux-Kernel Memory M INTRODUCTION ------------ -The Linux-kernel memory model (LKMM) is rather complex and obscure. -This is particularly evident if you read through the linux-kernel.bell -and linux-kernel.cat files that make up the formal version of the -memory model; they are extremely terse and their meanings are far from -clear. +The Linux-kernel memory consistency model (LKMM) is rather complex and +obscure. This is particularly evident if you read through the +linux-kernel.bell and linux-kernel.cat files that make up the formal +version of the model; they are extremely terse and their meanings are +far from clear. This document describes the ideas underlying the LKMM. It is meant -for people who want to understand how the memory model was designed. -It does not go into the details of the code in the .bell and .cat -files; rather, it explains in English what the code expresses -symbolically. +for people who want to understand how the model was designed. It does +not go into the details of the code in the .bell and .cat files; +rather, it explains in English what the code expresses symbolically. Sections 2 (BACKGROUND) through 5 (ORDERING AND CYCLES) are aimed -toward beginners; they explain what memory models are and the basic -notions shared by all such models. People already familiar with these -concepts can skim or skip over them. Sections 6 (EVENTS) through 12 -(THE FROM_READS RELATION) describe the fundamental relations used in -many memory models. Starting in Section 13 (AN OPERATIONAL MODEL), -the workings of the LKMM itself are covered. +toward beginners; they explain what memory consistency models are and +the basic notions shared by all such models. People already familiar +with these concepts can skim or skip over them. Sections 6 (EVENTS) +through 12 (THE FROM_READS RELATION) describe the fundamental +relations used in many models. Starting in Section 13 (AN OPERATIONAL +MODEL), the workings of the LKMM itself are covered. Warning: The code examples in this document are not written in the proper format for litmus tests. They don't include a header line, the
tools/memory-model/Documentation/explanation.txt A-cumulative; they only affect the propa · line 829
Quoted 6 lines
tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +1
-2@@ -827,8 +826,8 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -smp_read_barrier_depends(), rcu_read_lock(), rcu_read_unlock(), and -synchronize_rcu() fences have other properties which we discuss later. +read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have
rcu_read_lock()
Quoted 44 lines
tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +9
-7+other properties which we discuss later. PROPAGATION ORDER RELATION: cumul-fence@@ -988,8 +987,8 @@ Another possibility, not mentioned earli section, is: X and Y are both loads, X ->addr Y (i.e., there is an address - dependency from X to Y), and an smp_read_barrier_depends() - fence occurs between them. + dependency from X to Y), and X is a READ_ONCE() or an atomic + access. Dependencies can also cause instructions to be executed in program order. This is uncontroversial when the second instruction is a@@ -1015,9 +1014,9 @@ After all, a CPU cannot ask the memory s a particular location before it knows what that location is. However, the split-cache design used by Alpha can cause it to behave in a way that looks as if the loads were executed out of order (see the next -section for more details). For this reason, the LKMM does not include -address dependencies between read events in the ppo relation unless an -smp_read_barrier_depends() fence is present. +section for more details). The kernel includes a workaround for this +problem when the loads come from READ_ONCE(), and therefore the LKMM +includes address dependencies to loads in the ppo relation. On the other hand, dependencies can indirectly affect the ordering of two loads. This happens when there is a dependency from a load to a@@ -1114,11 +1113,12 @@ code such as the following: int *r1; int r2; - r1 = READ_ONCE(ptr); + r1 = ptr; r2 = READ_ONCE(*r1); } -can malfunction on Alpha systems. It is quite possible that r1 = &x +can malfunction on Alpha systems (notice that P1 uses an ordinary load +to read ptr instead of READ_ONCE()). It is quite possible that r1 = &x and r2 = 0 at the end, in spite of the address dependency. At first glance this doesn't seem to make sense. We know that the
tools/memory-model/Documentation/explanation.txt This could not have happened if the loca · line 1141
Quoted 1 line
tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +0
-0@@ -1141,11 +1141,15 @@ This could not have happened if the loca incoming stores in FIFO order. In constrast, other architectures
contrast
Quoted 89 lines
tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +17
-12maintain at least the appearance of FIFO order. -In practice, this difficulty is solved by inserting an -smp_read_barrier_depends() fence between P1's two loads. The effect -of this fence is to cause the CPU not to execute any po-later -instructions until after the local cache has finished processing all -the stores it has already received. Thus, if the code was changed to: +In practice, this difficulty is solved by inserting a special fence +between P1's two loads when the kernel is compiled for the Alpha +architecture. In fact, as of version 4.15, the kernel automatically +adds this fence (called smp_read_barrier_depends() and defined as +nothing at all on non-Alpha builds) after every READ_ONCE() and atomic +load. The effect of the fence is to cause the CPU not to execute any +po-later instructions until after the local cache has finished +processing all the stores it has already received. Thus, if the code +was changed to: P1() {@@ -1153,13 +1157,15 @@ the stores it has already received. Thu int r2; r1 = READ_ONCE(ptr); - smp_read_barrier_depends(); r2 = READ_ONCE(*r1); } then we would never get r1 = &x and r2 = 0. By the time P1 executed its second load, the x = 1 store would already be fully processed by -the local cache and available for satisfying the read request. +the local cache and available for satisfying the read request. Thus +we have yet another reason why shared data should always be read with +READ_ONCE() or another synchronization primitive rather than accessed +directly. The LKMM requires that smp_rmb(), acquire fences, and strong fences share this property with smp_read_barrier_depends(): They do not allow@@ -1751,11 +1757,10 @@ no further involvement from the CPU. Si the value of x, there is nothing for the smp_rmb() fence to act on. The LKMM defines a few extra synchronization operations in terms of -things we have already covered. In particular, rcu_dereference() and -lockless_dereference() are both treated as a READ_ONCE() followed by -smp_read_barrier_depends() -- which also happens to be how they are -defined in include/linux/rcupdate.h and include/linux/compiler.h, -respectively. +things we have already covered. In particular, rcu_dereference() is +treated as READ_ONCE() and rcu_assign_pointer() is treated as +smp_store_release() -- which is basically how the Linux kernel treats +them. There are a few oddball fences which need special treatment: smp_mb__before_atomic(), smp_mb__after_atomic(), andIndex: usb-4.x/tools/memory-model/linux-kernel.bell
===================================================================tools/memory-model/linux-kernel.bell renamed from tools/memory-model.orig/linux-kernel.bell +0
-1@@ -24,7 +24,6 @@ instructions RMW[{'once,'acquire,'releas enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || - 'rb_dep (*smp_read_barrier_depends*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) ||Index: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0
-2@@ -13,14 +13,12 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -lockless_dereference(X) __load{lderef}(X) rcu_dereference(X) __load{deref}(X) // Fences smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } -smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after-atomic} ; } smp_mb__after_spinlock() { __fence{after-spinlock} ; }Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================
tools/memory-model/Documentation/cheatsheet.txt line 7
Quoted 2 lines
tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +0
-0@@ -6,8 +6,7 @@ Store, e.g., WRITE_ONCE() Y Y Load, e.g., READ_ONCE() Y Y Y
READ_ONCE() now includes smp_read_barrier_depends, so this row should be divided to:
Simple Load Y Y Y
tools/memory-model/Documentation/cheatsheet.txt line 9
Quoted 2 lines
tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +0
-1Unsuccessful RMW operation Y Y Y -smp_read_barrier_depends() Y Y Y
READ_ONCE() Y Y Y Y > -*_dereference() Y Y Y Y > +rcu_dereference() Y Y Y Y > Successful *_acquire() R Y Y Y Y Y Y > Successful *_release() C Y Y Y W Y
Quoted 3 lines
tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +0
-0smp_rmb() Y R Y Y R
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 7 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.
Quoted 6 lines
Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Signed-off-by Alan Stern <[email protected]>
I queued this, but would welcome an update that addressed Akira's feedback as appropriate.
Thanx, Paul
Quoted 232 lines
---
Index: usb-4.x/tools/memory-model/linux-kernel.cat
===================================================================tools/memory-model/linux-kernel.cat renamed from tools/memory-model.orig/linux-kernel.cat +2
-5@@ -25,7 +25,6 @@ include "lock.cat" (*******************) (* Fences *) -let rb-dep = [R] ; fencerel(Rb_dep) ; [R] let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn] let wmb = [W] ; fencerel(Wmb) ; [W] let mb = ([M] ; fencerel(Mb) ; [M]) |@@ -61,11 +60,9 @@ let dep = addr | data let rwdep = (dep | ctrl) ; [W] let overwrite = co | fr let to-w = rwdep | (overwrite & int) -let rrdep = addr | (dep ; rfi) -let strong-rrdep = rrdep+ & rb-dep -let to-r = strong-rrdep | rfi-rel-acq +let to-r = addr | (dep ; rfi) | rfi-rel-acq let fence = strong-fence | wmb | po-rel | rmb | acq-po -let ppo = rrdep* ; (to-r | to-w | fence) +let ppo = to-r | to-w | fence (* Propagation: Ordering from release operations and strong fences. *) let A-cumul(r) = rfe? ; rIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt renamed from tools/memory-model.orig/Documentation/explanation.txt +43
-38@@ -1,5 +1,5 @@ -Explanation of the Linux-Kernel Memory Model -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Explanation of the Linux-Kernel Memory Consistency Model +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ :Author: Alan Stern <[email protected]> :Created: October 2017@@ -35,25 +35,24 @@ Explanation of the Linux-Kernel Memory M INTRODUCTION ------------ -The Linux-kernel memory model (LKMM) is rather complex and obscure. -This is particularly evident if you read through the linux-kernel.bell -and linux-kernel.cat files that make up the formal version of the -memory model; they are extremely terse and their meanings are far from -clear. +The Linux-kernel memory consistency model (LKMM) is rather complex and +obscure. This is particularly evident if you read through the +linux-kernel.bell and linux-kernel.cat files that make up the formal +version of the model; they are extremely terse and their meanings are +far from clear. This document describes the ideas underlying the LKMM. It is meant -for people who want to understand how the memory model was designed. -It does not go into the details of the code in the .bell and .cat -files; rather, it explains in English what the code expresses -symbolically. +for people who want to understand how the model was designed. It does +not go into the details of the code in the .bell and .cat files; +rather, it explains in English what the code expresses symbolically. Sections 2 (BACKGROUND) through 5 (ORDERING AND CYCLES) are aimed -toward beginners; they explain what memory models are and the basic -notions shared by all such models. People already familiar with these -concepts can skim or skip over them. Sections 6 (EVENTS) through 12 -(THE FROM_READS RELATION) describe the fundamental relations used in -many memory models. Starting in Section 13 (AN OPERATIONAL MODEL), -the workings of the LKMM itself are covered. +toward beginners; they explain what memory consistency models are and +the basic notions shared by all such models. People already familiar +with these concepts can skim or skip over them. Sections 6 (EVENTS) +through 12 (THE FROM_READS RELATION) describe the fundamental +relations used in many models. Starting in Section 13 (AN OPERATIONAL +MODEL), the workings of the LKMM itself are covered. Warning: The code examples in this document are not written in the proper format for litmus tests. They don't include a header line, the@@ -827,8 +826,8 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -smp_read_barrier_depends(), rcu_read_lock(), rcu_read_unlock(), and -synchronize_rcu() fences have other properties which we discuss later. +read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +other properties which we discuss later. PROPAGATION ORDER RELATION: cumul-fence@@ -988,8 +987,8 @@ Another possibility, not mentioned earli section, is: X and Y are both loads, X ->addr Y (i.e., there is an address - dependency from X to Y), and an smp_read_barrier_depends() - fence occurs between them. + dependency from X to Y), and X is a READ_ONCE() or an atomic + access. Dependencies can also cause instructions to be executed in program order. This is uncontroversial when the second instruction is a@@ -1015,9 +1014,9 @@ After all, a CPU cannot ask the memory s a particular location before it knows what that location is. However, the split-cache design used by Alpha can cause it to behave in a way that looks as if the loads were executed out of order (see the next -section for more details). For this reason, the LKMM does not include -address dependencies between read events in the ppo relation unless an -smp_read_barrier_depends() fence is present. +section for more details). The kernel includes a workaround for this +problem when the loads come from READ_ONCE(), and therefore the LKMM +includes address dependencies to loads in the ppo relation. On the other hand, dependencies can indirectly affect the ordering of two loads. This happens when there is a dependency from a load to a@@ -1114,11 +1113,12 @@ code such as the following: int *r1; int r2; - r1 = READ_ONCE(ptr); + r1 = ptr; r2 = READ_ONCE(*r1); } -can malfunction on Alpha systems. It is quite possible that r1 = &x +can malfunction on Alpha systems (notice that P1 uses an ordinary load +to read ptr instead of READ_ONCE()). It is quite possible that r1 = &x and r2 = 0 at the end, in spite of the address dependency. At first glance this doesn't seem to make sense. We know that the@@ -1141,11 +1141,15 @@ This could not have happened if the loca incoming stores in FIFO order. In constrast, other architectures maintain at least the appearance of FIFO order. -In practice, this difficulty is solved by inserting an -smp_read_barrier_depends() fence between P1's two loads. The effect -of this fence is to cause the CPU not to execute any po-later -instructions until after the local cache has finished processing all -the stores it has already received. Thus, if the code was changed to: +In practice, this difficulty is solved by inserting a special fence +between P1's two loads when the kernel is compiled for the Alpha +architecture. In fact, as of version 4.15, the kernel automatically +adds this fence (called smp_read_barrier_depends() and defined as +nothing at all on non-Alpha builds) after every READ_ONCE() and atomic +load. The effect of the fence is to cause the CPU not to execute any +po-later instructions until after the local cache has finished +processing all the stores it has already received. Thus, if the code +was changed to: P1() {@@ -1153,13 +1157,15 @@ the stores it has already received. Thu int r2; r1 = READ_ONCE(ptr); - smp_read_barrier_depends(); r2 = READ_ONCE(*r1); } then we would never get r1 = &x and r2 = 0. By the time P1 executed its second load, the x = 1 store would already be fully processed by -the local cache and available for satisfying the read request. +the local cache and available for satisfying the read request. Thus +we have yet another reason why shared data should always be read with +READ_ONCE() or another synchronization primitive rather than accessed +directly. The LKMM requires that smp_rmb(), acquire fences, and strong fences share this property with smp_read_barrier_depends(): They do not allow@@ -1751,11 +1757,10 @@ no further involvement from the CPU. Si the value of x, there is nothing for the smp_rmb() fence to act on. The LKMM defines a few extra synchronization operations in terms of -things we have already covered. In particular, rcu_dereference() and -lockless_dereference() are both treated as a READ_ONCE() followed by -smp_read_barrier_depends() -- which also happens to be how they are -defined in include/linux/rcupdate.h and include/linux/compiler.h, -respectively. +things we have already covered. In particular, rcu_dereference() is +treated as READ_ONCE() and rcu_assign_pointer() is treated as +smp_store_release() -- which is basically how the Linux kernel treats +them. There are a few oddball fences which need special treatment: smp_mb__before_atomic(), smp_mb__after_atomic(), andIndex: usb-4.x/tools/memory-model/linux-kernel.bell
===================================================================tools/memory-model/linux-kernel.bell renamed from tools/memory-model.orig/linux-kernel.bell +0
-1@@ -24,7 +24,6 @@ instructions RMW[{'once,'acquire,'releas enum Barriers = 'wmb (*smp_wmb*) || 'rmb (*smp_rmb*) || 'mb (*smp_mb*) || - 'rb_dep (*smp_read_barrier_depends*) || 'rcu-lock (*rcu_read_lock*) || 'rcu-unlock (*rcu_read_unlock*) || 'sync-rcu (*synchronize_rcu*) ||Index: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0
-2@@ -13,14 +13,12 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -lockless_dereference(X) __load{lderef}(X) rcu_dereference(X) __load{deref}(X) // Fences smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } -smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after-atomic} ; } smp_mb__after_spinlock() { __fence{after-spinlock} ; }Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +1
-2@@ -6,8 +6,7 @@ Store, e.g., WRITE_ONCE() Y Y Load, e.g., READ_ONCE() Y Y Y Unsuccessful RMW operation Y Y Y -smp_read_barrier_depends() Y Y Y -*_dereference() Y Y Y Y +rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Y smp_rmb() Y R Y Y R
Alan Stern
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 13 lines · Paul E. McKenney
Quoted 12 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Quoted 4 lines
Quoted 1 line
Signed-off-by Alan Stern <[email protected]>I queued this, but would welcome an update that addressed Akira's feedback as appropriate.
Is it too late to send a v2 of this patch? I didn't want to do it before now because the issue raised by Andrea wasn't settled. (Some could claim that it still isn't fully settled...)
Alan
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 15 lines · Alan Stern
Quoted 13 lines · Paul E. McKenney
Quoted 12 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Signed-off-by Alan Stern <[email protected]>I queued this, but would welcome an update that addressed Akira's feedback as appropriate.
Is it too late to send a v2 of this patch? I didn't want to do it before now because the issue raised by Andrea wasn't settled. (Some could claim that it still isn't fully settled...)
Would you be willing to send a delta patch? I can then place it directly on top of your original patch, and once it settles out, I can ask Ingo if he is willing to update the patch in -tip.
Thanx, Paul
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 21 lines · Paul E. McKenney
Quoted 20 lines · Alan Stern
Quoted 17 lines · Paul E. McKenney
Quoted 7 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.Quoted 6 lines
Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Signed-off-by Alan Stern <[email protected]>I queued this, but would welcome an update that addressed Akira's feedback as appropriate.
Quoted 7 lines
Quoted 3 lines
Is it too late to send a v2 of this patch? I didn't want to do it before now because the issue raised by Andrea wasn't settled. (Some could claim that it still isn't fully settled...)
Would you be willing to send a delta patch? I can then place it directly on top of your original patch, and once it settles out, I can ask Ingo if he is willing to update the patch in -tip.
It would be better to have followup fixes as separate patches. I applied the current set of fixes/improvements today to help move things along - it's all advancing very nicely!
Thanks,
Ingo
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 28 lines · Ingo Molnar
Quoted 25 lines · Paul E. McKenney
Quoted 23 lines · Alan Stern
Quoted 13 lines · Paul E. McKenney
Quoted 12 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Quoted 4 lines
Quoted 1 line
Signed-off-by Alan Stern <[email protected]>I queued this, but would welcome an update that addressed Akira's feedback as appropriate.
Is it too late to send a v2 of this patch? I didn't want to do it before now because the issue raised by Andrea wasn't settled. (Some could claim that it still isn't fully settled...)
Quoted 7 lines
Quoted 3 lines
Would you be willing to send a delta patch? I can then place it directly on top of your original patch, and once it settles out, I can ask Ingo if he is willing to update the patch in -tip.
It would be better to have followup fixes as separate patches. I applied the current set of fixes/improvements today to help move things along - it's all advancing very nicely!
Will do, and glad you like it!
Thanx, Paul
Alan Stern
maintainerPATCH 1/1 tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Paul E. McKenneyCommit bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
Andrea pointed out that the macro for rcu_dereference() in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit 5a8897cc7631 ("locking/atomics/alpha: Add smp_read_barrier_depends() to _release()/_relaxed() atomics") as an important precursor, and he contributed commit cb13b424e986 ("locking/xchg/alpha: Add unconditional memory barrier to cmpxchg()"), another prerequisite.
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)
tools/memory-model/Documentation/cheatsheet.txt +3 -3
@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Y
tools/memory-model/Documentation/explanation.txt +2 -2
@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.
@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fence
tools/memory-model/linux-kernel.def +1 -1
@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 17 lines · Alan Stern
Commit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
Andrea pointed out that the macro for
rcu_dereference()in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.
Quoted 4 lines
Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")
Thanks,
Andrea
Quoted 63 lines
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W YIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt +2
-2@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fenceIndex: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Paul E. McKenney
AppliedRe: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 18 lines · Andrea Parri
Quoted 17 lines · Alan Stern
Commit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
Andrea pointed out that the macro for
rcu_dereference()in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.
Quoted 6 lines
Quoted 4 lines
Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")Acked-by Andrea Parri <[email protected]>
Applied, thank you!
Thanx, Paul
Quoted 68 lines
Thanks,
AndreaQuoted 63 lines
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W YIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt +2
-2@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fenceIndex: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Paul E. McKenney
AppliedRe: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 17 lines · Alan Stern
Commit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
Andrea pointed out that the macro for
rcu_dereference()in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.
Quoted 4 lines
Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")
Queued, thank you!
Thanx, Paul
Quoted 62 lines
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W YIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt +2
-2@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fenceIndex: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 8 lines · Alan Stern
Commit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.
Quoted 9 lines
Andrea pointed out that the macro for
rcu_dereference()in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.
Quoted 5 lines
Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")
The change itself looks good to me.
Thanks, Akira
Quoted 62 lines
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W YIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt +2
-2@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fenceIndex: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 6 lines · Akira Yokosawa
Quoted 5 lines · Alan Stern
Commit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.
Quoted 8 lines
Quoted 3 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.
Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate READ_ONCE() implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies.
, if Alan and Andrea are OK with the amendment.
Also, please append my Acked-by.
Thanks, Akira
Quoted 84 lines
Quoted 9 lines
Andrea pointed out that the macro for
rcu_dereference()in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.Quoted 5 lines
Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")The change itself looks good to me.
Acked-by Akira Yokosawa <[email protected]>Thanks, Akira
Quoted 62 lines
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W YIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt +2
-2@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fenceIndex: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 24 lines · Akira Yokosawa
Quoted 6 lines · Akira Yokosawa
Quoted 5 lines · Alan Stern
Commit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.Quoted 8 lines
Quoted 3 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate
READ_ONCE()implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies.
Quoted 5 lines
, if Alan and Andrea are OK with the amendment.
Also, please append my Acked-by.
Acked-by Akira Yokosawa <[email protected]>
I can still amend this, and have added your Acked-by. If Alan and Andrea OK with your change, I will apply that also.
Thanx, Paul
Quoted 84 lines
Quoted 83 lines
Quoted 8 lines
Andrea pointed out that the macro for
rcu_dereference()in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.Quoted 5 lines
Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")The change itself looks good to me.
Acked-by Akira Yokosawa <[email protected]>Thanks, Akira
Quoted 62 lines
---
tools/memory-model/Documentation/cheatsheet.txt | 6 +++--- tools/memory-model/Documentation/explanation.txt | 4 ++-- tools/memory-model/linux-kernel.def | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W YIndex:
usb-4.x/tools/memory-model/Documentation/explanation.txt
===================================================================tools/memory-model/Documentation/explanation.txt +2
-2@@ -826,7 +826,7 @@ A-cumulative; they only affect the propa executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fenceIndex: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Andrea Parri
maintainerRe: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 3 lines · Paul E. McKenney
Quoted 2 lines · Akira Yokosawa
Quoted 1 line · Akira Yokosawa
On 2018/02/22 2:15, Alan Stern wrote:
[...]
Quoted 18 lines
Quoted 18 lines
Quoted 9 lines
Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate
READ_ONCE()implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies.
Quoted 8 lines
Quoted 5 lines
, if Alan and Andrea are OK with the amendment.
Also, please append my Acked-by.
Acked-by Akira Yokosawa <[email protected]>I can still amend this, and have added your Acked-by. If Alan and Andrea OK with your change, I will apply that also.
LGTM. Thanks,
Andrea
Quoted 1 line
Alan Stern
maintainerRe: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 27 lines · Andrea Parri
Quoted 3 lines · Paul E. McKenney
Quoted 2 lines · Akira Yokosawa
Quoted 1 line · Akira Yokosawa
On 2018/02/22 2:15, Alan Stern wrote:
[...]
Quoted 20 lines
Quoted 20 lines
Quoted 9 lines
Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate
READ_ONCE()implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies., if Alan and Andrea are OK with the amendment.
Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
Also, please append my Acked-by.
Acked-by Akira Yokosawa <[email protected]>I can still amend this, and have added your Acked-by. If Alan and Andrea OK with your change, I will apply that also.
LGTM. Thanks,
Me too.
Alan
Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 31 lines · Alan Stern
Quoted 29 lines · Andrea Parri
Quoted 3 lines · Paul E. McKenney
Quoted 2 lines · Akira Yokosawa
Quoted 1 line · Akira Yokosawa
On 2018/02/22 2:15, Alan Stern wrote:
[...]
Quoted 22 lines
Quoted 22 lines
Quoted 9 lines
Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate
READ_ONCE()implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies., if Alan and Andrea are OK with the amendment.
Also, please append my Acked-by.
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
Acked-by Akira Yokosawa <[email protected]>I can still amend this, and have added your Acked-by. If Alan and Andrea OK with your change, I will apply that also.
LGTM. Thanks,
Me too.
Very good, how about this for the new version?
Thanx, Paul
------------------------------------------------------------------------
commit 21ede43970e50b7397420f17ed08bb02c187e2eb
Author: Alan Stern <[email protected]> Date: Wed Feb 21 12:15:56 2018 -0500
tools/memory-model: Update: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Commit bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should indicate that READ_ONCE() now implies an address dependency. Andrea suggested documenting the relationship betwwen unsuccessful RMW operations and address dependencies.
Andrea pointed out that the macro for rcu_dereference() in linux.def should now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit 5a8897cc7631 ("locking/atomics/alpha: Add smp_read_barrier_depends() to _release()/_relaxed() atomics") as an important precursor, and he contributed commit cb13b424e986 ("locking/xchg/alpha: Add unconditional memory barrier to cmpxchg()"), another prerequisite.
tools/memory-model/Documentation/cheatsheet.txt +3 -3
@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Y
tools/memory-model/Documentation/explanation.txt +2 -2
@@ -826,7 +826,7 @@ A-cumulative; they only affect the propagation of stores that are executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have other properties which we discuss later.
@@ -1138,7 +1138,7 @@ final effect is that even though the two loads really are executed in program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fence
tools/memory-model/linux-kernel.def +1 -1
@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Akira Yokosawa
reviewerRe: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 58 lines · Paul E. McKenney
Quoted 31 lines · Alan Stern
Quoted 29 lines · Andrea Parri
Quoted 3 lines · Paul E. McKenney
Quoted 2 lines · Akira Yokosawa
Quoted 1 line · Akira Yokosawa
On 2018/02/22 2:15, Alan Stern wrote:
[...]
Quoted 22 lines
Quoted 22 lines
Quoted 9 lines
Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate
READ_ONCE()implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies., if Alan and Andrea are OK with the amendment.
Also, please append my Acked-by.
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
Acked-by Akira Yokosawa <[email protected]>I can still amend this, and have added your Acked-by. If Alan and Andrea OK with your change, I will apply that also.
LGTM. Thanks,
Me too.
Very good, how about this for the new version?
Thanx, Paul
------------------------------------------------------------------------
commit
21ede43970e50b7397420f17ed08bb02c187e2ebAuthor: Alan Stern <[email protected]> Date: Wed Feb 21 12:15:56 2018 -0500
tools/memory-model: Update: Remove rb-dep,smp_read_barrier_depends, andlockless_dereferenceCommit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.
Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should indicate that
READ_ONCE()now implies an address dependency. Andrea suggested documenting the relationship betwwen unsuccessful RMW operations and address dependencies.
Looks good. But I've found a remaining typo in the patch. See below.
Quoted 40 lines
Quoted 1 line
Andrea pointed out that the macro for
rcu_dereference()in linux.defshould now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit
5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")Acked-by Andrea Parri <[email protected]>Acked-by Akira Yokosawa <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Ytools/memory-model/Documentation/explanation.txt +0
-0
tools/memory-model/Documentation/explanation.txt A-cumulative; they only affect the propagation of stores that are · line 829
Quoted 5 lines
tools/memory-model/Documentation/explanation.txt +1
-1@@ -826,7 +826,7 @@ A-cumulative; they only affect the propagation of stores that are executed on C before the fence (i.e., those which precede the fence in program order). -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have
rcu_read_lock()
Don't ask why I missed this in the first place...
Paul, can you fix this directly?
Thanks, Akira
Quoted 26 lines
tools/memory-model/Documentation/explanation.txt +1
-1other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two loads really are executed in program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fencetools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 109 lines · Akira Yokosawa
Quoted 58 lines · Paul E. McKenney
Quoted 31 lines · Alan Stern
Quoted 29 lines · Andrea Parri
Quoted 3 lines · Paul E. McKenney
Quoted 2 lines · Akira Yokosawa
Quoted 1 line · Akira Yokosawa
On 2018/02/22 2:15, Alan Stern wrote:
[...]
Quoted 22 lines
Quoted 22 lines
Quoted 9 lines
Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate how unsuccessful RMW operations relate to address dependencies.
My point was to separate unannotated loads from
READ_ONCE(), if the cheatsheet should concern such accesses as well. Unsuccessful RMW operations were brought up by Andrea.Paul, can you amend above paragraph in the change log to something like:
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should be updated to indicate
READ_ONCE()implies address dependency, which invited Andrea's observation that it should also be updated to indicate how unsuccessful RMW operations relate to address dependencies., if Alan and Andrea are OK with the amendment.
Also, please append my Acked-by.
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
Acked-by Akira Yokosawa <[email protected]>I can still amend this, and have added your Acked-by. If Alan and Andrea OK with your change, I will apply that also.
LGTM. Thanks,
Me too.
Very good, how about this for the new version?
Thanx, Paul
------------------------------------------------------------------------
commit
21ede43970e50b7397420f17ed08bb02c187e2ebAuthor: Alan Stern <[email protected]> Date: Wed Feb 21 12:15:56 2018 -0500
tools/memory-model: Update: Remove rb-dep,smp_read_barrier_depends, andlockless_dereferenceCommit
bf28ae562744("tools/memory-model: Remove rb-dep,smp_read_barrier_depends, andlockless_dereference") was accidentally merged too early, while it was still in RFC form. This patch adds in the missing pieces.Quoted 4 lines
Akira pointed out some typos in the original patch, and he noted that cheatsheet.txt should indicate that
READ_ONCE()now implies an address dependency. Andrea suggested documenting the relationship betwwen unsuccessful RMW operations and address dependencies.Looks good. But I've found a remaining typo in the patch. See below.
Quoted 43 lines
Quoted 1 line
Andrea pointed out that the macro for
rcu_dereference()in linux.defshould now use the "once" annotation instead of "deref". He also suggested that the comments should mention commit
5a8897cc7631("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics") as an important precursor, and he contributed commitcb13b424e986("locking/xchg/alpha: Add unconditional memory barrier tocmpxchg()"), another prerequisite.Signed-off-by Alan Stern <[email protected]>Suggested-by Akira Yokosawa <[email protected]>Suggested-by Andrea Parri <[email protected]>Fixes bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")Acked-by Andrea Parri <[email protected]>Acked-by Akira Yokosawa <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>tools/memory-model/Documentation/cheatsheet.txt +3
-3@@ -1,11 +1,11 @@ Prior Operation Subsequent Operation --------------- --------------------------- C Self R W RWM Self R W DR DW RMW SV - __ ---- - - --- ---- - - -- -- --- -- + -- ---- - - --- ---- - - -- -- --- -- Store, e.g., WRITE_ONCE() Y Y -Load, e.g., READ_ONCE() Y Y Y -Unsuccessful RMW operation Y Y Y +Load, e.g., READ_ONCE() Y Y Y Y +Unsuccessful RMW operation Y Y Y Y rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Ytools/memory-model/Documentation/explanation.txt +0
-0@@ -826,7 +826,7 @@ A-cumulative; they only affect the propagation of stores that are executed on C before the fence (i.e., those which precede the fence in program order).
Quoted 8 lines
tools/memory-model/Documentation/explanation.txt A-cumulative; they only affect the propagation of stores that are · line 829
Quoted 2 lines
tools/memory-model/Documentation/explanation.txt +1
-1-read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences havercu_read_lock()Don't ask why I missed this in the first place...
Paul, can you fix this directly?
Done!
Thanx, Paul
Quoted 29 lines
Thanks, Akira
Quoted 26 lines
tools/memory-model/Documentation/explanation.txt +1
-1other properties which we discuss later.@@ -1138,7 +1138,7 @@ final effect is that even though the two loads really are executed in program order, it appears that they aren't. This could not have happened if the local cache had processed the -incoming stores in FIFO order. In constrast, other architectures +incoming stores in FIFO order. By contrast, other architectures maintain at least the appearance of FIFO order. In practice, this difficulty is solved by inserting a special fencetools/memory-model/linux-kernel.def +1
-1@@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -rcu_dereference(X) __load{deref}(X) +rcu_dereference(X) __load{once}(X) // Fences smp_mb() { __fence{mb} ; }
Andrea Parri
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 15 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Signed-off-by Alan Stern <[email protected]>---
[...]
tools/memory-model/linux-kernel.def WRITE_ONCE(X,V) { __store{once}(X,V); } · line 16
Quoted 10 lines
Index: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0
-1@@ -13,14 +13,12 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -lockless_dereference(X) __load{lderef}(X) rcu_dereference(X) __load{deref}(X)
^^^ __load{once}
Quoted 14 lines
tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0
-1// Fences smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } -smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after-atomic} ; } smp_mb__after_spinlock() { __fence{after-spinlock} ; }Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================
tools/memory-model/Documentation/cheatsheet.txt line 12
Quoted 9 lines
tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +1
-2@@ -6,8 +6,7 @@ Store, e.g., WRITE_ONCE() Y Y Load, e.g., READ_ONCE() Y Y Y Unsuccessful RMW operation Y Y Y -smp_read_barrier_depends() Y Y Y -*_dereference() Y Y Y Y +rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Y smp_rmb() Y R Y Y R
Akira's observation about READ_ONCE extends to all (annotated) loads. In fact, it also applies to loads corresponding to unsuccessful RMW operations; consider, for example, the following variation of MP+onceassign+derefonce:
C T
{
y=z;
z=0;
}
P0(int *x, int **y)
{
WRITE_ONCE(*x, 1);
smp_store_release(y, x);
}P1(int **y, int *z)
{
int *r0;
int r1;
r0 = cmpxchg_relaxed(y, z, z);
r1 = READ_ONCE(*r0);
}exists (1:r0=x /\ 1:r1=0)
The final state is allowed w/o the patch, and forbidden w/ the patch.
This also reminds me of
5a8897cc7631fa544d079c443800f4420d1b173f
("locking/atomics/alpha: Add smp_read_barrier_depends() to _release()/_relaxed() atomics")
(that we probably want to mention in the commit message).
Andrea
Quoted 2 lines
Andrea Parri
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 86 lines · Andrea Parri
Quoted 15 lines · Alan Stern
Since commit
76ebbe78f739("locking/barriers: Add implicitsmp_read_barrier_depends()toREAD_ONCE()") was merged for the 4.15 kernel, it has not been necessary to usesmp_read_barrier_depends(). Similarly, commit59ecbbe7b31c("locking/barriers: Killlockless_dereference()") removedlockless_dereference()from the kernel.Since these primitives are no longer part of the kernel, they do not belong in the Linux Kernel Memory Consistency Model. This patch removes them, along with the internal rb-dep relation, and updates the revelant documentation.
Signed-off-by Alan Stern <[email protected]>---
[...]
tools/memory-model/linux-kernel.def WRITE_ONCE(X,V) { __store{once}(X,V); } · line 16
Quoted 10 lines
Index: usb-4.x/tools/memory-model/linux-kernel.def
===================================================================tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0
-1@@ -13,14 +13,12 @@ WRITE_ONCE(X,V) { __store{once}(X,V); } smp_store_release(X,V) { __store{release}(*X,V); } smp_load_acquire(X) __load{acquire}(*X) rcu_assign_pointer(X,V) { __store{release}(X,V); } -lockless_dereference(X) __load{lderef}(X) rcu_dereference(X) __load{deref}(X)^^^ __load{once}
Quoted 14 lines
tools/memory-model/linux-kernel.def renamed from tools/memory-model.orig/linux-kernel.def +0
-1// Fences smp_mb() { __fence{mb} ; } smp_rmb() { __fence{rmb} ; } smp_wmb() { __fence{wmb} ; } -smp_read_barrier_depends() { __fence{rb_dep}; } smp_mb__before_atomic() { __fence{before-atomic} ; } smp_mb__after_atomic() { __fence{after-atomic} ; } smp_mb__after_spinlock() { __fence{after-spinlock} ; }Index:
usb-4.x/tools/memory-model/Documentation/cheatsheet.txt
===================================================================tools/memory-model/Documentation/cheatsheet.txt line 12
Quoted 9 lines
tools/memory-model/Documentation/cheatsheet.txt renamed from tools/memory-model.orig/Documentation/cheatsheet.txt +1
-2@@ -6,8 +6,7 @@ Store, e.g., WRITE_ONCE() Y Y Load, e.g., READ_ONCE() Y Y Y Unsuccessful RMW operation Y Y Y -smp_read_barrier_depends() Y Y Y -*_dereference() Y Y Y Y +rcu_dereference() Y Y Y Y Successful *_acquire() R Y Y Y Y Y Y Successful *_release() C Y Y Y W Y smp_rmb() Y R Y Y RAkira's observation about
READ_ONCEextends to all (annotated) loads. In fact, it also applies to loads corresponding to unsuccessful RMW operations; consider, for example, the following variation of MP+onceassign+derefonce:C T
{
y=z;
z=0;
}
P0(int *x, int **y){ WRITE_ONCE(*x, 1); smp_store_release(y, x); }
P1(int **y, int *z){ int *r0; int r1; r0 = cmpxchg_relaxed(y, z, z); r1 = READ_ONCE(*r0); }exists (1:r0=x /\ 1:r1=0)
Quoted 8 lines
The final state is allowed w/o the patch, and forbidden w/ the patch.
This also reminds me of
5a8897cc7631fa544d079c443800f4420d1b173f
("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics")(that we probably want to mention in the commit message).
Please also notice that 5a8897cc7631f only touched alpha's atomic.h:
I see no corresponding commit/change on {,cmp}xchg.h (where the "mb" is currently conditionally executed).
Andrea
Quoted 6 lines
Andrea
Quoted 2 lines
Alan Stern
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 32 lines · Andrea Parri
Quoted 32 lines
Akira's observation about
READ_ONCEextends to all (annotated) loads. In fact, it also applies to loads corresponding to unsuccessful RMW operations; consider, for example, the following variation of MP+onceassign+derefonce:C T
{
y=z;
z=0;
}
P0(int *x, int **y){ WRITE_ONCE(*x, 1); smp_store_release(y, x); }
P1(int **y, int *z){ int *r0; int r1; r0 = cmpxchg_relaxed(y, z, z); r1 = READ_ONCE(*r0); }exists (1:r0=x /\ 1:r1=0)
The final state is allowed w/o the patch, and forbidden w/ the patch.
This also reminds me of
Quoted 8 lines
Quoted 4 lines
5a8897cc7631fa544d079c443800f4420d1b173f
("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics")(that we probably want to mention in the commit message).
Please also notice that
5a8897cc7631fonly touched alpha'satomic.h:
I see no corresponding commit/change on {,cmp}xchg.h (where the "mb" is currently conditionally executed).
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
Note that operations like atomic_add_unless() already include memory barriers.
Alan
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 39 lines · Alan Stern
Quoted 37 lines · Andrea Parri
Quoted 36 lines
Akira's observation about
READ_ONCEextends to all (annotated) loads. In fact, it also applies to loads corresponding to unsuccessful RMW operations; consider, for example, the following variation of MP+onceassign+derefonce:C T
{
y=z;
z=0;
}
P0(int *x, int **y){ WRITE_ONCE(*x, 1); smp_store_release(y, x); }
P1(int **y, int *z){ int *r0; int r1; r0 = cmpxchg_relaxed(y, z, z); r1 = READ_ONCE(*r0); }exists (1:r0=x /\ 1:r1=0)
The final state is allowed w/o the patch, and forbidden w/ the patch.
This also reminds me of
5a8897cc7631fa544d079c443800f4420d1b173f
("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics")(that we probably want to mention in the commit message).
Quoted 8 lines
Quoted 3 lines
Please also notice that
5a8897cc7631fonly touched alpha'satomic.h:
I see no corresponding commit/change on {,cmp}xchg.h (where the "mb" is currently conditionally executed).This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
I vote for adding the barrier on Alpha. However, I don't know of any code in the Linux kernel that relies on read-to-read address dependency ordering headed by a failing RMW operation, so I don't feel all that strongly about this.
Quoted 2 lines
Note that operations like
atomic_add_unless()already include memory barriers.
And I don't see an atomic_add_unless_relaxed(), so we are good on this one. So far, anyway! ;-)
Thanx, Paul
Peter Zijlstra
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 10 lines · Paul E. McKenney
Quoted 4 lines · Alan Stern
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
I vote for adding the barrier on Alpha. However, I don't know of any code in the Linux kernel that relies on read-to-read address dependency ordering headed by a failing RMW operation, so I don't feel all that strongly about this.
Right, but not knowing doesn't mean doesn't exist, and most certainly doesn't mean will never exist.
Quoted 5 lines
Quoted 2 lines
Note that operations like
atomic_add_unless()already include memory barriers.And I don't see an
atomic_add_unless_relaxed(), so we are good on this one. So far, anyway! ;-)
Not the point, add_unless() is a conditional operation, and therefore doesn't need to imply anything when failing.
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 7 lines · Peter Zijlstra
Quoted 6 lines · Paul E. McKenney
Quoted 4 lines · Alan Stern
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
Quoted 7 lines
Quoted 4 lines
I vote for adding the barrier on Alpha. However, I don't know of any code in the Linux kernel that relies on read-to-read address dependency ordering headed by a failing RMW operation, so I don't feel all that strongly about this.
Right, but not knowing doesn't mean doesn't exist, and most certainly doesn't mean will never exist.
Fair enough, safety first!
Quoted 8 lines
Quoted 5 lines
Quoted 2 lines
Note that operations like
atomic_add_unless()already include memory barriers.And I don't see an
atomic_add_unless_relaxed(), so we are good on this one. So far, anyway! ;-)Not the point,
add_unless()is a conditional operation, and therefore doesn't need to imply anything when failing.
Plus it doesn't return a pointer, so there is no problem with dereferences. Unless someone wants to use its return value as an array index and rely on dependency ordering to the array, but I would NAK that use case.
Thanx, Paul
Alan Stern
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 24 lines · Paul E. McKenney
Quoted 7 lines · Peter Zijlstra
Quoted 6 lines · Paul E. McKenney
Quoted 4 lines · Alan Stern
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
Quoted 7 lines
Quoted 4 lines
I vote for adding the barrier on Alpha. However, I don't know of any code in the Linux kernel that relies on read-to-read address dependency ordering headed by a failing RMW operation, so I don't feel all that strongly about this.
Right, but not knowing doesn't mean doesn't exist, and most certainly doesn't mean will never exist.
Fair enough, safety first!
Quoted 6 lines
Quoted 5 lines
Quoted 2 lines
Note that operations like
atomic_add_unless()already include memory barriers.And I don't see an
atomic_add_unless_relaxed(), so we are good on this one. So far, anyway! ;-)
Quoted 6 lines
Quoted 2 lines
Not the point,
add_unless()is a conditional operation, and therefore doesn't need to imply anything when failing.Plus it doesn't return a pointer, so there is no problem with dereferences. Unless someone wants to use its return value as an array index and rely on dependency ordering to the array, but I would NAK that use case.
You may not get the chance to NAK it.
We need to be consistent. Array indexing is indeed a form of address dependency, so either we assert that the dependency is enforced when the array index is derived from a failed atomic operation, or else we assert that failed atomic operations do not create address dependencies.
Alan
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 33 lines · Alan Stern
Quoted 24 lines · Paul E. McKenney
Quoted 7 lines · Peter Zijlstra
Quoted 6 lines · Paul E. McKenney
Quoted 4 lines · Alan Stern
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
Quoted 7 lines
Quoted 4 lines
I vote for adding the barrier on Alpha. However, I don't know of any code in the Linux kernel that relies on read-to-read address dependency ordering headed by a failing RMW operation, so I don't feel all that strongly about this.
Right, but not knowing doesn't mean doesn't exist, and most certainly doesn't mean will never exist.
Fair enough, safety first!
Quoted 6 lines
Quoted 5 lines
Quoted 2 lines
Note that operations like
atomic_add_unless()already include memory barriers.And I don't see an
atomic_add_unless_relaxed(), so we are good on this one. So far, anyway! ;-)Quoted 6 lines
Quoted 2 lines
Not the point,
add_unless()is a conditional operation, and therefore doesn't need to imply anything when failing.Plus it doesn't return a pointer, so there is no problem with dereferences. Unless someone wants to use its return value as an array index and rely on dependency ordering to the array, but I would NAK that use case.
Quoted 7 lines
You may not get the chance to NAK it.
We need to be consistent. Array indexing is indeed a form of address dependency, so either we assert that the dependency is enforced when the array index is derived from a failed atomic operation, or else we assert that failed atomic operations do not create address dependencies.
The problem is that the compiler can get up to a great deal more mischief with integers than it can with pointers. I agree that it would be quite challenging to represent this distinction between integers and pointers in the model, but then again herd does not yet support array indexing anyway.
So for the time being, anyway, I must stand by my NAK.
Thanx, Paul
Andrea Parri
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 39 lines · Alan Stern
Quoted 37 lines · Andrea Parri
Quoted 36 lines
Akira's observation about
READ_ONCEextends to all (annotated) loads. In fact, it also applies to loads corresponding to unsuccessful RMW operations; consider, for example, the following variation of MP+onceassign+derefonce:C T
{
y=z;
z=0;
}
P0(int *x, int **y){ WRITE_ONCE(*x, 1); smp_store_release(y, x); }
P1(int **y, int *z){ int *r0; int r1; r0 = cmpxchg_relaxed(y, z, z); r1 = READ_ONCE(*r0); }exists (1:r0=x /\ 1:r1=0)
The final state is allowed w/o the patch, and forbidden w/ the patch.
This also reminds me of
5a8897cc7631fa544d079c443800f4420d1b173f
("locking/atomics/alpha: Addsmp_read_barrier_depends()to _release()/_relaxed() atomics")(that we probably want to mention in the commit message).
Quoted 8 lines
Quoted 3 lines
Please also notice that
5a8897cc7631fonly touched alpha'satomic.h:
I see no corresponding commit/change on {,cmp}xchg.h (where the "mb" is currently conditionally executed).This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
I'd like to continue to treat R[once] and R*[once] equally if possible. Given the (unconditional) smp_read_barrier_depends in READ_ONCE and in atomics, it seems reasonable to have it unconditionally in cmpxchg.
As with the following patch?
Andrea
---
arch/alpha/include/asm/xchg.h +7 -8
@@ -128,10 +128,9 @@ ____xchg(, volatile void *ptr, unsigned long x, int size)* store NEW in MEM. Return the initial value in MEM. Success is * indicated by comparing RETURN with OLD. * -* The memory barrier should be placed in SMP only when we actually -* make the change. If we don't change anything (so if the returned -* prev is equal to old) then we aren't acquiring anything new and -* we don't need any memory barrier as far I can tell. +* The memory barrier is placed in SMP unconditionally, in order to +* guarantee that dependency ordering is preserved when a dependency +* is headed by an unsuccessful operation. */ static inline unsigned long
@@ -150,8 +149,8 @@ ____cmpxchg(_u8, volatile char *m, unsigned char old, unsigned char new)" or %1,%2,%2\n" " stq_c %2,0(%4)\n" " beq %2,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous"
@@ -177,8 +176,8 @@ ____cmpxchg(_u16, volatile short *m, unsigned short old, unsigned short new)" or %1,%2,%2\n" " stq_c %2,0(%4)\n" " beq %2,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous"
@@ -200,8 +199,8 @@ ____cmpxchg(_u32, volatile int *m, int old, int new)" mov %4,%1\n" " stl_c %1,%2\n" " beq %1,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous"
@@ -223,8 +222,8 @@ ____cmpxchg(_u64, volatile long *m, unsigned long old, unsigned long new)" mov %4,%1\n" " stq_c %1,%2\n" " beq %1,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous"
Quoted 6 lines
Note that operations like
atomic_add_unless()already include memory barriers.Alan
Peter Zijlstra
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 58 lines · Andrea Parri
I'd like to continue to treat R[once] and R*[once] equally if possible. Given the (unconditional)
smp_read_barrier_dependsinREAD_ONCEand in atomics, it seems reasonable to have it unconditionally in cmpxchg.As with the following patch?
Andrea
---
arch/alpha/include/asm/xchg.h +6
-7@@ -128,10 +128,9 @@ ____xchg(, volatile void *ptr, unsigned long x, int size)* store NEW in MEM. Return the initial value in MEM. Success is * indicated by comparing RETURN with OLD. * -* The memory barrier should be placed in SMP only when we actually -* make the change. If we don't change anything (so if the returned -* prev is equal to old) then we aren't acquiring anything new and -* we don't need any memory barrier as far I can tell. +* The memory barrier is placed in SMP unconditionally, in order to +* guarantee that dependency ordering is preserved when a dependency +* is headed by an unsuccessful operation. */ static inline unsigned long@@ -150,8 +149,8 @@ ____cmpxchg(_u8, volatile char *m, unsigned char old, unsigned char new)" or %1,%2,%2\n" " stq_c %2,0(%4)\n" " beq %2,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous" @@ -177,8 +176,8 @@ ____cmpxchg(_u16, volatile short *m, unsigned short old, unsigned short new)" or %1,%2,%2\n" " stq_c %2,0(%4)\n" " beq %2,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous" @@ -200,8 +199,8 @@ ____cmpxchg(_u32, volatile int *m, int old, int new)" mov %4,%1\n" " stl_c %1,%2\n" " beq %1,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous"
arch/alpha/include/asm/xchg.h ____cmpxchg(_u64, volatile long *m, unsigned long old, unsigned long new) · line 229
Quoted 9 lines
arch/alpha/include/asm/xchg.h +1
-1@@ -223,8 +222,8 @@ ____cmpxchg(_u64, volatile long *m, unsigned long old, unsigned long new)" mov %4,%1\n" " stq_c %1,%2\n" " beq %1,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous"
ACK
Alan Stern
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Andrea ParriQuoted 10 lines · Andrea Parri
Quoted 4 lines
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
I'd like to continue to treat R[once] and R*[once] equally if possible. Given the (unconditional)
smp_read_barrier_dependsinREAD_ONCEand in atomics, it seems reasonable to have it unconditionally in cmpxchg.As with the following patch?
Yes, this seems reasonable to me. If Will gives it his "Acked-by" to go with Peter's, you should submit it to Ingo Molnar.
And once this is made, there shouldn't be any trouble with the proposed patch for the memory model.
Alan
Quoted 68 lines
Andrea
---
arch/alpha/include/asm/xchg.h +7
-8@@ -128,10 +128,9 @@ ____xchg(, volatile void *ptr, unsigned long x, int size)* store NEW in MEM. Return the initial value in MEM. Success is * indicated by comparing RETURN with OLD. * -* The memory barrier should be placed in SMP only when we actually -* make the change. If we don't change anything (so if the returned -* prev is equal to old) then we aren't acquiring anything new and -* we don't need any memory barrier as far I can tell. +* The memory barrier is placed in SMP unconditionally, in order to +* guarantee that dependency ordering is preserved when a dependency +* is headed by an unsuccessful operation. */ static inline unsigned long@@ -150,8 +149,8 @@ ____cmpxchg(_u8, volatile char *m, unsigned char old, unsigned char new)" or %1,%2,%2\n" " stq_c %2,0(%4)\n" " beq %2,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous" @@ -177,8 +176,8 @@ ____cmpxchg(_u16, volatile short *m, unsigned short old, unsigned short new)" or %1,%2,%2\n" " stq_c %2,0(%4)\n" " beq %2,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous" @@ -200,8 +199,8 @@ ____cmpxchg(_u32, volatile int *m, int old, int new)" mov %4,%1\n" " stl_c %1,%2\n" " beq %1,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous" @@ -223,8 +222,8 @@ ____cmpxchg(_u64, volatile long *m, unsigned long old, unsigned long new)" mov %4,%1\n" " stq_c %1,%2\n" " beq %1,3f\n" - __ASM__MB"2:\n" + __ASM__MB".subsection 2\n" "3: br 1b\n" ".previous" Quoted 7 lines
Note that operations like
atomic_add_unless()already include memory barriers.Alan
Peter Zijlstra
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 4 lines · Alan Stern
This leaves us with a question: Do we want to change the kernel by adding memory barriers after unsuccessful RMW operations on Alpha, or do we want to change the model by excluding such operations from address dependencies?
I think we want to put all the 'pain and weirdness' in Alpha.
Peter Zijlstra
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
In reply to Alan SternQuoted 2 lines · Alan Stern
Note that operations like
atomic_add_unless()already include memory barriers.
It is valid for atomic_add_unless() to not imply any barriers when the addition doesn't happen.
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 6 lines · Peter Zijlstra
Quoted 2 lines · Alan Stern
Note that operations like
atomic_add_unless()already include memory barriers.It is valid for
atomic_add_unless()to not imply any barriers when the addition doesn't happen.
Agreed, given that atomic_add_unless() just returns 0 or 1, not the pointer being added. Of course, the __atomic_add_unless() function that it calls is another story, as it does return the old value. Sigh. And __atomic_add_unless() is called directly from some code. All of which looks to be counters rather than pointers, thankfully.
So, do we want to rely on atomic_add_unless() always being invoked on counters rather than pointers, or does it need an smp_read_barrier_depends()?
Thanx, Paul
Andrea Parri
maintainerRe: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 14 lines · Paul E. McKenney
Quoted 6 lines · Peter Zijlstra
Quoted 2 lines · Alan Stern
Note that operations like
atomic_add_unless()already include memory barriers.It is valid for
atomic_add_unless()to not imply any barriers when the addition doesn't happen.Agreed, given that
atomic_add_unless()just returns 0 or 1, not the pointer being added. Of course, the__atomic_add_unless()function that it calls is another story, as it does return the old value. Sigh. And__atomic_add_unless()is called directly from some code. All of which looks to be counters rather than pointers, thankfully.
Quoted 3 lines
So, do we want to rely on
atomic_add_unless()always being invoked on counters rather than pointers, or does it need ansmp_read_barrier_depends()?
alpha's implementation of __atomic_add_unless() has an unconditional smp_mb() before returning so, as far as dependencies are concerned, these seem fine.
Andrea
Quoted 3 lines
Thanx, Paul
Re: [PATCH] tools/memory-model: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
Quoted 15 lines · Andrea Parri
Quoted 14 lines · Paul E. McKenney
Quoted 6 lines · Peter Zijlstra
Quoted 2 lines · Alan Stern
Note that operations like
atomic_add_unless()already include memory barriers.It is valid for
atomic_add_unless()to not imply any barriers when the addition doesn't happen.Agreed, given that
atomic_add_unless()just returns 0 or 1, not the pointer being added. Of course, the__atomic_add_unless()function that it calls is another story, as it does return the old value. Sigh. And__atomic_add_unless()is called directly from some code. All of which looks to be counters rather than pointers, thankfully.
Quoted 6 lines
Quoted 3 lines
So, do we want to rely on
atomic_add_unless()always being invoked on counters rather than pointers, or does it need ansmp_read_barrier_depends()?alpha's implementation of
__atomic_add_unless()has an unconditionalsmp_mb()before returning so, as far as dependencies are concerned, these seem fine.
Very good!
Thanx, Paul
Andrea Parri
maintainerRe: Trial of conflict resolution of Alan's patch
In reply to Paul E. McKenneyQuoted 90 lines · Paul E. McKenney
Quoted 49 lines · Alan Stern
Quoted 10 lines · Akira Yokosawa
So, I attempted to rebase the patch to current (somewhat old) master of
https://github.com/aparri/memory-model. Why? Because the lkmm branch in Paul's -rcu tree doesn't have linux-kernel-hardware.cat.However, after this change, Z6.0+pooncelock+pooncelock+pombonce still has the result "Sometimes". I must have done something wrong in the conflict resolution.
Note: I have almost no idea what this patch is doing. I'm just hoping to give a starting point of a discussion.
Yes, that litmus test gives "Sometimes" both with and without the patch. But consider instead this slightly changed version of that test, in which P2 reads Z instead of writing it:
C Z6.0-variant
{}
P0(int *x, int *y, spinlock_t *mylock){ spin_lock(mylock); WRITE_ONCE(*x, 1); WRITE_ONCE(*y, 1); spin_unlock(mylock); }
P1(int *y, int *z, spinlock_t *mylock){ int r0; spin_lock(mylock); r0 = READ_ONCE(*y); WRITE_ONCE(*z, 1); spin_unlock(mylock); }
P2(int *x, int *z){ int r1; int r2; r2 = READ_ONCE(*z); smp_mb(); r1 = READ_ONCE(*x); }Quoted 7 lines
exists (1:r0=1 /\ 2:r2=1 /\ 2:r1=0)
Without the patch, this test gives "Sometimes"; with the patch it gives "Never". That is what I thought Paul was talking about originally.
Sorry if my misunderstanding caused too much confusion for other people.
Ah, I did indeed get confused. I have changed the "Result:" for Z6.0+pooncelock+pooncelock+pombonce.litmus back to "Never", as in the patch below (which I merged into the patch adding all the comments).
I have added the above test as ISA2+pooncelock+pooncelock+pombonce.litmus, with the Result: of Sometimes with you (Alan) as author and with your Signed-off-by -- please let me know if you would prefer some other approach.
Please change the Result: when sending the proposed patch. Or please let me know if you would like me to apply the forward-port that Akira sent, in which case I will add the Result: change to that patch. Or for that matter, Akira might repost his forward-port of your patch with this change.
Thanx, Paul
------------------------------------------------------------------------
commit
b2950420e1154131c0667f1ac58666bad3a06a69Author: Paul E. McKenney <[email protected]> Date: Thu Feb 15 10:35:25 2018 -0800fixup! EXP
litmus_tests: Add comments explaining tests' purposesSigned-off-by Paul E. McKenney <[email protected]>tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +0
-0
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus line 4
Quoted 5 lines
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +1
-1@@ -1,7 +1,7 @@ C Z6.0+pooncelock+pooncelock+pombonce (* - * Result: Never + * Result: Somtimes
nit: s/Somtimes/Sometimes
Andrea
Quoted 4 lines
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +0
-0* * This example demonstrates that a pair of accesses made by different * processes each while holding a given lock will not necessarily be
Re: Trial of conflict resolution of Alan's patch
Quoted 93 lines · Andrea Parri
Quoted 92 lines · Paul E. McKenney
Quoted 49 lines · Alan Stern
Quoted 10 lines · Akira Yokosawa
So, I attempted to rebase the patch to current (somewhat old) master of
https://github.com/aparri/memory-model. Why? Because the lkmm branch in Paul's -rcu tree doesn't have linux-kernel-hardware.cat.However, after this change, Z6.0+pooncelock+pooncelock+pombonce still has the result "Sometimes". I must have done something wrong in the conflict resolution.
Note: I have almost no idea what this patch is doing. I'm just hoping to give a starting point of a discussion.
Yes, that litmus test gives "Sometimes" both with and without the patch. But consider instead this slightly changed version of that test, in which P2 reads Z instead of writing it:
C Z6.0-variant
{}
P0(int *x, int *y, spinlock_t *mylock){ spin_lock(mylock); WRITE_ONCE(*x, 1); WRITE_ONCE(*y, 1); spin_unlock(mylock); }
P1(int *y, int *z, spinlock_t *mylock){ int r0; spin_lock(mylock); r0 = READ_ONCE(*y); WRITE_ONCE(*z, 1); spin_unlock(mylock); }
P2(int *x, int *z){ int r1; int r2; r2 = READ_ONCE(*z); smp_mb(); r1 = READ_ONCE(*x); }Quoted 7 lines
exists (1:r0=1 /\ 2:r2=1 /\ 2:r1=0)
Without the patch, this test gives "Sometimes"; with the patch it gives "Never". That is what I thought Paul was talking about originally.
Sorry if my misunderstanding caused too much confusion for other people.
Ah, I did indeed get confused. I have changed the "Result:" for Z6.0+pooncelock+pooncelock+pombonce.litmus back to "Never", as in the patch below (which I merged into the patch adding all the comments).
I have added the above test as ISA2+pooncelock+pooncelock+pombonce.litmus, with the Result: of Sometimes with you (Alan) as author and with your Signed-off-by -- please let me know if you would prefer some other approach.
Please change the Result: when sending the proposed patch. Or please let me know if you would like me to apply the forward-port that Akira sent, in which case I will add the Result: change to that patch. Or for that matter, Akira might repost his forward-port of your patch with this change.
Thanx, Paul
------------------------------------------------------------------------
commit
b2950420e1154131c0667f1ac58666bad3a06a69Author: Paul E. McKenney <[email protected]> Date: Thu Feb 15 10:35:25 2018 -0800fixup! EXP
litmus_tests: Add comments explaining tests' purposesSigned-off-by Paul E. McKenney <[email protected]>tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +0
-0@@ -1,7 +1,7 @@ C Z6.0+pooncelock+pooncelock+pombonce
Quoted 5 lines
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus line 4
Quoted 3 lines
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +1
-1(* - * Result: Never + * Result: Somtimesnit: s/Somtimes/Sometimes
Good catch, fixed!
Thanx, Paul
Quoted 8 lines
Andrea
Quoted 4 lines
tools/memory-model/litmus-tests/Z6.0+pooncelock+pooncelock+pombonce.litmus +0
-0* * This example demonstrates that a pair of accesses made by different * processes each while holding a given lock will not necessarily be
Akira Yokosawa
reviewerRe: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
In reply to Paul E. McKenneyQuoted 41 lines · Paul E. McKenney
Quoted 26 lines · Akira Yokosawa
Quoted 25 lines · Paul E. McKenney
Quoted 23 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Quoted 8 lines
Quoted 5 lines
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
So, herdtools7.7.48 is now available by "opam update; opam upgrade". Maybe mentioning the required version of herdtools7 in README would help.
Or have some way for the memory model's .cat files to state what version they need, but in the meantime please see the patch below. But even with such version specification, we probably want to have the version in the README...
Quoted 8 lines
Quoted 3 lines
One glitch I'm aware of is one of the output of klitmus7 fails to compile on kernels prior to 4.14, because of
smp_mb__after_spinlock()used in Z6.0+pooncelock+poonceLock+pombonce.litmus.This is one advantage of having the memory model in the kernel source itself -- the versions match. And people can always fire up a different kernel version (for example, within a VM) to run the output of klitmus7.
There is another unfortunate mismatch in kernel and herdtools7 updates.
klitmus7 in herdtools7 7.48 requires definition of ACCESS_ONCE() in kernel headers, but it has been removed in Linux 4.15. This means klitmus7 of herdtools7 7.48 works only on Linux 4.14.
In the repository of herdtools7, with the suggestion of Andrea, commit e87d7f9287d1 ("klitmus: Use WRITE_ONCE and READ_ONCE in place of deprecated ACCESS_ONCE in "user" synchronization barrier") has addressed the issue.
This series is intended to be merged in Linux 4.17 merge window, and hopefully we can have another update of herdtools7 by the time the merge window opens...
Dependency to out-of-tree tools looks quite tricky. We need some neat way to manage things.
Umm...
Thanks, Akira
[...]
Re: [PATCH v2] tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")
Quoted 61 lines · Akira Yokosawa
Quoted 41 lines · Paul E. McKenney
Quoted 26 lines · Akira Yokosawa
Quoted 25 lines · Paul E. McKenney
Quoted 23 lines · Akira Yokosawa
Quoted 1 line
From
7c1f497a9a51e8db1a94c8a7ef0b74b235aaab88Mon Sep 17 00:00:00 2001From: Akira Yokosawa <[email protected]>
Date: Fri, 9 Feb 2018 04:51:05 -0800
Subject: [PATCH v2]tools/memory-model: Make compat with herd7 7.47 ("-" -> "_")As of herd7 7.47, these '-'s are not permitted and end up in errors such as:
File "./linux-kernel.def", line 44, characters 29-30:
unexpected '-' (in macros)Partial revert of commit
2d5fba7782d6("linux-kernel*: Make RCU identifiers match ASPLOS paper") in the repository at
https://github.com/aparri/memory-model can restore the compatibility with herd7 7.47.Reported-by Patrick Bellasi <[email protected]>Suggested-by Andrea Parri <[email protected]>Signed-off-by Akira Yokosawa <[email protected]>---
Paul,FWIW, this is a squashed version relative to patch 07/10 in the RFC series.
Quoted 8 lines
Quoted 5 lines
Thank you, Akira!
I am going to hold off on this for a bit to see if we can instead get a new release of herd7, but if we can't. this might well be a very good way to go.
So, herdtools7.7.48 is now available by "opam update; opam upgrade". Maybe mentioning the required version of herdtools7 in README would help.
Or have some way for the memory model's .cat files to state what version they need, but in the meantime please see the patch below. But even with such version specification, we probably want to have the version in the README...
Quoted 8 lines
Quoted 3 lines
One glitch I'm aware of is one of the output of klitmus7 fails to compile on kernels prior to 4.14, because of
smp_mb__after_spinlock()used in Z6.0+pooncelock+poonceLock+pombonce.litmus.This is one advantage of having the memory model in the kernel source itself -- the versions match. And people can always fire up a different kernel version (for example, within a VM) to run the output of klitmus7.
There is another unfortunate mismatch in kernel and herdtools7 updates.
klitmus7 in herdtools7 7.48 requires definition of
ACCESS_ONCE()in kernel headers, but it has been removed in Linux 4.15. This means klitmus7 of herdtools7 7.48 works only on Linux 4.14.In the repository of herdtools7, with the suggestion of Andrea, commit
e87d7f9287d1("klitmus: UseWRITE_ONCEandREAD_ONCEin place of deprecatedACCESS_ONCEin "user" synchronization barrier") has addressed the issue.
Quoted 8 lines
This series is intended to be merged in Linux 4.17 merge window, and hopefully we can have another update of herdtools7 by the time the merge window opens...
Dependency to out-of-tree tools looks quite tricky. We need some neat way to manage things.
Umm...
Yes, dependencies on out-of-tree tools is quite tricky, witness the "fun and excitement" provided even by GCC from time to time.
I agree that another herdtools7 release is required, and I do very much thank you for testing this and finding this issue!
Thanx, Paul