[5.7,092/204] perf intel-pt: Fix duplicate branch after CBR

Message ID	20200820091610.928636133@linuxfoundation.org
State	Superseded
Headers	show Return-Path: <SRS0=nT5k=B6=vger.kernel.org=stable-owner@kernel.org> From: Greg Kroah-Hartman <gregkh@linuxfoundation.org> To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>, stable@vger.kernel.org, Adrian Hunter <adrian.hunter@intel.com>, Andi Kleen <ak@linux.intel.com>, Arnaldo Carvalho de Melo <acme@redhat.com>, Jiri Olsa <jolsa@redhat.com> Subject: [PATCH 5.7 092/204] perf intel-pt: Fix duplicate branch after CBR Date: Thu, 20 Aug 2020 11:19:49 +0200 Message-Id: <20200820091610.928636133@linuxfoundation.org> In-Reply-To: <20200820091606.194320503@linuxfoundation.org> References: <20200820091606.194320503@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Sender: stable-owner@vger.kernel.org Precedence: bulk
Series	None \| expand [5.7,003/204] genirq/PM: Always unlock IRQ descriptor in rearm_wake_irq() [5.7,004/204] PCI: hotplug: ACPI: Fix context refcounting in acpiphp_grab_context() [5.7,007/204] PCI: Add device even if driver attach failed [5.7,010/204] btrfs: allow use of global block reserve for balance item deletion [5.7,013/204] btrfs: ref-verify: fix memory leak in add_block_entry [5.7,015/204] btrfs: stop incremening log_batch for the log root tree when syncing log [5.7,016/204] btrfs: only commit delayed items at fsync if we are logging a directory [5.7,019/204] btrfs: pass checksum type via BTRFS_IOC_FS_INFO ioctl [5.7,021/204] btrfs: move the chunk_mutex in btrfs_read_chunk_tree [5.7,022/204] btrfs: relocation: review the call sites which can be interrupted by signal [5.7,023/204] btrfs: add missing check for nocow and compression inode flags [5.7,025/204] btrfs: return EROFS for BTRFS_FS_STATE_ERROR cases [5.7,026/204] btrfs: sysfs: use NOFS for device creation [5.7,027/204] btrfs: dont WARN if we abort a transaction with EROFS [5.7,028/204] btrfs: fix race between page release and a fast fsync [5.7,029/204] btrfs: dont show full path of bind mounts in subvol= [5.7,031/204] btrfs: only search for left_info if there is no right_info in try_merge_free_space [5.7,032/204] btrfs: inode: fix NULL pointer dereference if inode doesnt need compression [5.7,036/204] btrfs: fix return value mixup in btrfs_get_extent [5.7,038/204] dt-bindings: iio: io-channel-mux: Fix compatible string in example code [5.7,040/204] iio: imu: st_lsm6dsx: reset hw ts after resume [5.7,041/204] xtensa: add missing exclusive access state management [5.7,043/204] cifs: Fix leak when handling lease break for cached root fid [5.7,047/204] pinctrl: ingenic: Enhance support for IRQ_TYPE_EDGE_BOTH [5.7,049/204] media: venus: fix multiple encoder crash [5.7,052/204] net: stmmac: dwmac1000: provide multicast filter fallback [5.7,054/204] irqchip/gic-v4.1: Ensure accessing the correct RD when writing INVALLR [5.7,057/204] selftests/seccomp: Set NNP for TSYNC ESRCH flag test [5.7,058/204] md/raid5: Fix Force reconstruct-write io stuck in degraded raid5 [5.7,059/204] bcache: allocate meta data pages as compound pages [5.7,060/204] bcache: fix overflow in offset_to_stripe() [5.7,062/204] driver core: Avoid binding drivers to dead devices [5.7,064/204] MIPS: qi_lb60: Fix routing to audio amplifier [5.7,065/204] MIPS: SGI-IP27: always enable NUMA in Kconfig [5.7,068/204] mm/hugetlb: fix calculation of adjust_range_if_pmd_sharing_possible [5.7,070/204] hugetlbfs: remove call to huge_pte_alloc without i_mmap_rwsem [5.7,072/204] ocfs2: change slot number type s16 to u16 [5.7,073/204] mm/page_counter.c: fix protection usage propagation [5.7,076/204] ftrace: Setup correct FTRACE_FL_REGS flags for module [5.7,078/204] tracing/hwlat: Honor the tracing_cpumask [5.7,081/204] watchdog: f71808e_wdt: remove use of wrong watchdog_info option [5.7,083/204] ceph: set sec_context xattr on symlink creation [5.7,084/204] ceph: handle zero-length feature mask in session messages [5.7,088/204] bootconfig: Fix to find the initargs correctly [5.7,089/204] perf probe: Fix wrong variable warning when the probe point is not found [5.7,091/204] perf intel-pt: Fix FUP packet state [5.7,092/204] perf intel-pt: Fix duplicate branch after CBR [5.7,093/204] gfs2: Never call gfs2_block_zero_range with an open transaction [5.7,094/204] remoteproc: qcom: q6v5: Update running state before requesting stop [5.7,096/204] remoteproc: qcom_q6v5_mss: Validate modem blob firmware size before load [5.7,097/204] drm/imx: imx-ldb: Disable both channels for split mode in enc->disable() [5.7,099/204] crypto: algif_aead - Only wake up when ctx->more is zero [5.7,102/204] mtd: rawnand: fsl_upm: Remove unused mtd var [5.7,103/204] mtd: rawnand: brcmnand: ECC error handling on EDU transfers [5.7,106/204] RDMA/ipoib: Fix ABBA deadlock with ipoib_reap_ah() [5.7,107/204] rtc: cpcap: fix range [5.7,109/204] media: rockchip: rga: Introduce color fmt macros and refactor CSC mode logic [5.7,112/204] sched/uclamp: Protect uclamp fast path code with static key [5.7,115/204] selftests/bpf: test_progs use another shell exit on non-actions [5.7,116/204] selftests/bpf: test_progs avoid minus shell exit codes [5.7,119/204] USB: serial: ftdi_sio: fix break and sysrq handling [5.7,122/204] RDMA/counter: Only bind user QPs in auto mode [5.7,124/204] mmc: renesas_sdhi_internal_dmac: clean up the code for dma complete [5.7,125/204] rtw88: pci: disable aspm for platform inter-op with module parameter [5.7,126/204] rtc: pl031: fix set_alarm by adding back call to alarm_irq_enable [5.7,127/204] crypto: caam - Remove broken arc4 support [5.7,129/204] gpu: ipu-v3: image-convert: Wait for all EOFs before completing a tile [5.7,130/204] dm rq: dont call blk_mq_queue_stopped() in dm_stop_queue() [5.7,132/204] f2fs: compress: fix to avoid memory leak on cc->cpages [5.7,135/204] selftests/powerpc: ptrace-pkey: Update the test to mark an invalid pkey correctly [5.7,137/204] iommu/omap: Check for failure of a call to omap_iommu_dump_ctx [5.7,138/204] clk: qcom: gcc: fix sm8150 GPU and NPU clocks [5.7,140/204] iommu/vt-d: Enforce PASID devTLB field mask [5.7,142/204] f2fs: compress: fix to update isize when overwriting compressed file [5.7,145/204] nvme: fix deadlock in disconnect during scan_work and/or ana_work [5.7,147/204] pwm: bcm-iproc: handle clk_get_rate() return [5.7,148/204] tools build feature: Use CC and CXX from parent [5.7,150/204] nfs: ensure correct writeback errors are returned on close() [5.7,152/204] clk: bcm2835: Do not use prediv with bcm2711s PLLs [5.7,154/204] libnvdimm/security: ensure sysfs poll thread woke up and fetch updated attr [5.7,155/204] openrisc: Fix oops caused when dumping stack [5.7,157/204] nfs: nfs_file_write() should check for writeback errors [5.7,159/204] watchdog: rti-wdt: balance pm runtime enable calls [5.7,160/204] md-cluster: Fix potential error pointer dereference in resize_bitmaps() [5.7,161/204] x86/tsr: Fix tsc frequency enumeration bug on Lightning Mountain SoC [5.7,163/204] selftests/bpf: Prevent runqslower from racing on building bpftool [5.7,165/204] vdpa_sim: init iommu lock [5.7,170/204] s390/test_unwind: fix possible memleak in test_unwind() [5.7,171/204] s390/Kconfig: add missing ZCRYPT dependency to VFIO_AP [5.7,173/204] nfs: Fix getxattr kernel panic and memory overflow [5.7,176/204] fs/minix: set s_maxbytes correctly [5.7,178/204] fs/minix: remove expected error message in block_to_path() [5.7,180/204] test_kmod: avoid potential double free in trigger_config_run_type() [5.7,181/204] i2c: iproc: fix race between client unreg and isr [5.7,186/204] tools build feature: Quote CC and CXX for their arguments [5.7,191/204] ASoC: tegra: Add audio mclk parent configuration [5.7,193/204] drm/radeon: fix fb_div check in ni_init_smc_spll_table() [5.7,195/204] drm/panfrost: Use kvfree() to free bo->sgts [5.7,197/204] drm/amdgpu: fix ordering of psp suspend [5.7,198/204] drm/omap: force runtime PM suspend on system suspend [5.7,202/204] drm/amdgpu: Fix bug where DPM is not enabled after hibernate and resume [5.7,204/204] drm/amd/display: dchubbub p-state warning during surface planes switch

Message ID

20200820091610.928636133@linuxfoundation.org

State

Superseded

Headers

From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Adrian Hunter <adrian.hunter@intel.com>,
	Andi Kleen <ak@linux.intel.com>,
	Arnaldo Carvalho de Melo <acme@redhat.com>, Jiri Olsa <jolsa@redhat.com>
Subject: [PATCH 5.7 092/204] perf intel-pt: Fix duplicate branch after CBR
Date: Thu, 20 Aug 2020 11:19:49 +0200
Message-Id: <20200820091610.928636133@linuxfoundation.org>
In-Reply-To: <20200820091606.194320503@linuxfoundation.org>
References: <20200820091606.194320503@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Sender: stable-owner@vger.kernel.org
Precedence: bulk

Series

None | expand

Commit Message

Greg Kroah-Hartman Aug. 20, 2020, 9:19 a.m. UTC

From: Adrian Hunter <adrian.hunter@intel.com>

commit a58a057ce65b52125dd355b7d8b0d540ea267a5f upstream.

CBR events can result in a duplicate branch event, because the state
type defaults to a branch. Fix by clearing the state type.

Example: trace 'sleep' and hope for a frequency change

 Before:

   $ perf record -e intel_pt//u sleep 0.1
   [ perf record: Woken up 1 times to write data ]
   [ perf record: Captured and wrote 0.034 MB perf.data ]
   $ perf script --itrace=bpe > before.txt

 After:

   $ perf script --itrace=bpe > after.txt
   $ diff -u before.txt after.txt
#  --- before.txt  2020-07-07 14:42:18.191508098 +0300
#  +++ after.txt   2020-07-07 14:42:36.587891753 +0300
   @@ -29673,7 +29673,6 @@
               sleep 93431 [007] 15411.619905:          1  branches:u:                 0 [unknown] ([unknown]) =>     7f0818abb2e0 clock_nanosleep@@GLIBC_2.17+0x0 (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
               sleep 93431 [007] 15411.619905:          1  branches:u:      7f0818abb30c clock_nanosleep@@GLIBC_2.17+0x2c (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>                0 [unknown] ([unknown])
               sleep 93431 [007] 15411.720069:         cbr:  cbr: 15 freq: 1507 MHz ( 56%)         7f0818abb30c clock_nanosleep@@GLIBC_2.17+0x2c (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
   -           sleep 93431 [007] 15411.720069:          1  branches:u:      7f0818abb30c clock_nanosleep@@GLIBC_2.17+0x2c (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>                0 [unknown] ([unknown])
               sleep 93431 [007] 15411.720076:          1  branches:u:                 0 [unknown] ([unknown]) =>     7f0818abb30e clock_nanosleep@@GLIBC_2.17+0x2e (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
               sleep 93431 [007] 15411.720077:          1  branches:u:      7f0818abb323 clock_nanosleep@@GLIBC_2.17+0x43 (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>     7f0818ac0eb7 __nanosleep+0x17 (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
               sleep 93431 [007] 15411.720077:          1  branches:u:      7f0818ac0ebf __nanosleep+0x1f (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>     55cb7e4c2827 rpl_nanosleep+0x97 (/usr/bin/sleep)

Fixes: 91de8684f1cff ("perf intel-pt: Cater for CBR change in PSB+")
Fixes: abe5a1d3e4bee ("perf intel-pt: Decoder to output CBR changes immediately")
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: stable@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200710151104.15137-3-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -1977,8 +1977,10 @@  next:
 			 * possibility of another CBR change that gets caught up
 			 * in the PSB+.
 			 */
-			if (decoder->cbr != decoder->cbr_seen)
+			if (decoder->cbr != decoder->cbr_seen) {
+				decoder->state.type = 0;
 				return 0;
+			}
 			break;
 
 		case INTEL_PT_PIP:
@@ -2019,8 +2021,10 @@  next:
 
 		case INTEL_PT_CBR:
 			intel_pt_calc_cbr(decoder);
-			if (decoder->cbr != decoder->cbr_seen)
+			if (decoder->cbr != decoder->cbr_seen) {
+				decoder->state.type = 0;
 				return 0;
+			}
 			break;
 
 		case INTEL_PT_MODE_EXEC:

[5.7,092/204] perf intel-pt: Fix duplicate branch after CBR

Commit Message

Patch