[PATCH v4 00/11] perf: Refine barriers for AUX ring buffer

From: Leo Yan
Date: Sun Jul 11 2021 - 06:41:25 EST


This patch series is to refine the memory barriers for AUX ring buffer.

Patches 01 ~ 04 to address the barriers usage in the kernel. The first
patch is to make clear comment for how to use the barriers between the
data store and aux_head store, this asks the driver to make sure the
data is visible. Patches 02 ~ 04 is to refine the drivers for barriers
after the data store.

Patch 05 is to use WRITE_ONCE() for updating aux_tail.

Patches 06 ~ 09 is to drop the legacy __sync functions, and polish for
duplicate code and cleanup the build and feature test after
SYNC_COMPARE_AND_SWAP is not used.

Patch 10 introduces a new global variable to indicate the kernel runs in
64-bit mode which can be used to confirm if in compat mode; patch 11
introduces variant functions for accessing AUX head/tail, it resolves
the aotmicity for reading head pointer, and returns error for the tail
is bigger than 4GB.

Have testes the patches on Arm64 Juno platform.

Changes from v3:
- Removed the inapprocate paragraph in the commit log for patch "perf
auxtrace: Drop legacy __sync functions" (Adrian);
- Added new patch to remove feature-sync-compare-and-swap test (Adrian);
- Th patch for "perf auxtrace: Use WRITE_ONCE() for updating aux_tail",
is a standlone and simple change, so moved it ahead in the patch set
for better ordering;
- Minor improvement for commit logs in the last two patches.

Changes from v2:
- Removed auxtrace_mmap__read_snapshot_head(), which has the duplicated
code with auxtrace_mmap__read_head();
- Cleanuped the build for HAVE_SYNC_COMPARE_AND_SWAP_SUPPORT (Adrian);
- Added global variable "kernel_is_64_bit" (Adrian);
- Added compat variants compat_auxtrace_mmap__{read_head|write_tail}
(Adrian).


Leo Yan (11):
perf/ring_buffer: Add comment for barriers on AUX ring buffer
coresight: tmc-etr: Add barrier after updating AUX ring buffer
coresight: tmc-etf: Add comment for store ordering
perf/x86: Add barrier after updating bts
perf auxtrace: Use WRITE_ONCE() for updating aux_tail
perf auxtrace: Drop legacy __sync functions
perf auxtrace: Remove auxtrace_mmap__read_snapshot_head()
perf: Cleanup for HAVE_SYNC_COMPARE_AND_SWAP_SUPPORT
tools: Remove feature-sync-compare-and-swap feature detection
perf env: Set flag for kernel is 64-bit mode
perf auxtrace: Add compat_auxtrace_mmap__{read_head|write_tail}

arch/x86/events/intel/bts.c | 3 +
.../hwtracing/coresight/coresight-tmc-etf.c | 6 +
.../hwtracing/coresight/coresight-tmc-etr.c | 8 ++
kernel/events/ring_buffer.c | 9 ++
tools/build/Makefile.feature | 1 -
tools/build/feature/Makefile | 4 -
tools/build/feature/test-all.c | 4 -
.../feature/test-sync-compare-and-swap.c | 15 ---
tools/perf/Makefile.config | 4 -
tools/perf/util/auxtrace.c | 19 ++-
tools/perf/util/auxtrace.h | 109 ++++++++++++++----
tools/perf/util/env.c | 17 ++-
tools/perf/util/env.h | 1 +
13 files changed, 136 insertions(+), 64 deletions(-)
delete mode 100644 tools/build/feature/test-sync-compare-and-swap.c

--
2.25.1