CoreSight

coresight@lists.linaro.org

6 participants
2535 discussions

by Raj Swaminathan - C

Hello, I have had some success streaming instruction trace to my Linux PC using RDDI (from ARM Development studio 2019) and a DStreamer ST. However I am also having some problems: 1. Failure to stop trace cleanly I have enabled streaming from an ETM attached to a Cortex-R5 core. As described in the RDDI documentation, to stop streaming, I call StreamingTrace_Flush() and wait for RDDI_STREAMING_TRACE_EVENT_TYPE_END_OF_DATA event. Next i call StreamingTrace_Stop() which succeeds and then try StreamingTrace_Detach() which seems to fail with the following assert. tpp.c:84: __pthread_tpp_change_priority: Assertion `new_prio == -1 || (new_prio >= fifo_min_prio && new_prio <= fifo_max_prio)' failed. Aborted (core dumped) If i try to avoid calling StreamingTrace_{Detach, Stop} then Debug_CloseConn to the core never returns and i get a seg fault. 2. Raw trace does not start with A-SYNC bits I do see A-SYNC bits later but my understanding was ETM always sends the A-SYNC bits first. Cheers, raj

6 years, 1 month

[PATCH v5 0/3] coresight: etm4x: save/restore ETMv4 context across CPU low power states

by Andrew Murray

Some hardware will ignore bit TRCPDCR.PU which is used to signal to hardware that power should not be removed from the trace unit. Let's mitigate against this by conditionally saving and restoring the trace unit state when the CPU enters low power states. This patchset introduces a firmware property named 'arm,coresight-loses-context-with-cpu' - when this is present the hardware state will be conditionally saved and restored. A module parameter 'pm_save_enable' is also introduced which can be configured to override the firmware property. This parameter also provides a means to save/restore state when external agents are used. The hardware state is only ever saved and restored when a coresight session is present. The last patch should be considered as an RFC as further consideration is required relating to where the pm_save_enable parameter lives and to determine if the external agent support should be a pm_save_enable option or a new kernel option. Changes since v4: - Rename fwnode property to "arm,coresight-loses-context-with-cpu" as this doesn't imply a software policy - Update the device tree binding document to indicate that this property isn't specific to ETMs - also provide a longer description more generic description with an example of why it might be used - Set the module parameter at probe based on the value determined by firmware. The user can still override the firmware via the kernel command line, this has the effect of hiding the PARAM_PM_SAVE_FIRMWARE option from the user - though we still internally use it to allow us to determine if the user has set the parameter. - Remove unnecessary call to smp_processor_id - Move etm4_needs_save_restore helper to coresight.c and rename - Rebased onto coresight/next a04d8683f577 ("...ity of etm4_os_unlock comment") - Drop Reviewed-By from Suzuki on "coresight: etm4x: save/restore st..." patch as content changed too much - Add module option to that keeps clocks/power enabled at probe and saves state when external or self-hosted is in use. Changes since v3: - Only save/restore when self-hosted is being used and detect this without relying on the coresight registers (which may not be available) - Only allocate memory for etmv4_save_state at probe time when configuration indicates it may be used - Set pm_save_enable param to 0444 such that it is static after boot - Save/restore TRCPDCR - Add missing comments to struct etm4_drvdata documentation - Rebased onto coresight/next (8f1f9857) Changes since v2: - Move the PM notifier block from drvdata to file static - Add section names to document references - Add additional information to commit messages - Remove trcdvcvr and trcdvcmr from saved state and add a comment to describe why - Ensure TRCPDCR_PU is set after restore and add a comment to explain why we bother toggling TRCPDCR_PU on save/restore - Reword the pm_save_enable options and add comments - Miscellaneous style changes - Move device tree binding documentation to its own patch Changes since v1: - Rebased onto coresight/next - Correcly pass bit number rather than BIT macro to coresight_timeout - Abort saving state if a timeout occurs - Fix completely broken pm_notify handling and unregister handler on error - Use state_needs_restore to ensure state is restored only once - Add module parameter description to existing boot_enable parameter and use module_param instead of module_param_named - Add firmware bindings for coresight-needs-save-restore - Rename 'disable_pm_save' to 'pm_save_enable' which allows for disabled, enabled or firmware - Update comment on etm4_os_lock, it incorrectly indicated that the code unlocks the trace registers - Add comments to explain use of OS lock during save/restore - Fix incorrect error description whilst waiting for PM stable - Add WARN_ON_ONCE when cpu isn't as expected during save/restore - Various updates to commit messages Andrew Murray (3): coresight: etm4x: save/restore state across CPU low power states dt-bindings: arm: coresight: Add support for coresight-loses-context-with-cpu coresight: etm4x: save/restore state for external agents .../devicetree/bindings/arm/coresight.txt | 9 + drivers/hwtracing/coresight/coresight-etm4x.c | 339 +++++++++++++++++- drivers/hwtracing/coresight/coresight-etm4x.h | 64 ++++ drivers/hwtracing/coresight/coresight.c | 8 +- include/linux/coresight.h | 13 + 5 files changed, 431 insertions(+), 2 deletions(-) -- 2.21.0

6 years, 2 months

Tracing and decoding support for trace hardware without trace formatter

by Linu Cherian

Hi, We have an ARM v8 based SOC that has got Marvell's ARM trace implementation, with the Trace formatter disabled. ie. FFSR[FtNotPresent] in ETR register space reports 0b1. Have couple of queries related to the tracing and decoding of this hardware configuration. 1. Does openCSD tool supports decoding of trace capture from hardware that has trace formatter disabled. >From the openCSD RRADME, "The library will decode formatted trace in three stages: Frame Deformatting : Removal CoreSight frame formatting from individual trace streams. Packet Processing : Separate individual trace streams into discrete packets. Packet Decode : Convert the packets into fully decoded trace describing the program flow on a core." So, is it possible for openCSD to carry out decoding bypassing the first stage mentioned above. 2. Linux Coresight ETR driver as part of managing the trace buffer wrap condition, a barrier packet is written at the beginning which is essentially a Frame synchronization packet. This gives an impression that that ETR driver is making an assumption that the trace formatter is implemented in hardware. Should we not need a fix here, to accommodate hardware configurations that doesn't support trace formatter as well ? Thanks. Linu Cherian

6 years, 2 months

YOUR DHL SHIPMENT DELIVERY.

by DHL International GmbH

6 years, 2 months

[PATCH v5] perf machine: arm/arm64: Improve completeness for kernel address space

by Leo Yan

Arm and arm64 architecture reserve some memory regions prior to the symbol '_stext' and these memory regions later will be used by device module and BPF jit. The current code misses to consider these memory regions thus any address in the regions will be taken as user space mode, but perf cannot find the corresponding dso with the wrong CPU mode so we misses to generate samples for device module and BPF related trace data. This patch parse the link scripts to get the memory size prior to start address and reduce this size from 'machine>->kernel_start', then can get a fixed up kernel start address which contain memory regions for device module and BPF. Finally, machine__get_kernel_start() can reflect more complete kernel memory regions and perf can successfully generate samples. The reason for parsing the link scripts is Arm architecture changes text offset dependent on different platforms, which define multiple text offsets in $kernel/arch/arm/Makefile. This offset is decided when build kernel and the final value is extended in the link script, so we can extract the used value from the link script. We use the same way to parse arm64 link script as well. If fail to find the link script, the pre start memory size is assumed as zero, in this case it has no any change caused with this patch. Below is detailed info for testing this patch: - Install or build LLVM/Clang; - Configure perf with ~/.perfconfig: root@debian:~# cat ~/.perfconfig # this file is auto-generated. [llvm] clang-path = /mnt/build/llvm-build/build/install/bin/clang kbuild-dir = /mnt/linux-kernel/linux-cs-dev/ clang-opt = "-g" dump-obj = true [trace] show_zeros = yes show_duration = no no_inherit = yes show_timestamp = no show_arg_names = no args_alignment = 40 show_prefix = yes - Run 'perf trace' command with eBPF event: root@debian:~# perf trace -e string \ -e $kernel/tools/perf/examples/bpf/augmented_raw_syscalls.c - Read eBPF program memory mapping in kernel: root@debian:~# echo 1 > /proc/sys/net/core/bpf_jit_kallsyms root@debian:~# cat /proc/kallsyms | grep -E "bpf_prog_.+_sys_[enter|exit]" ffff00000008a0d0 t bpf_prog_e470211b846088d5_sys_enter [bpf] ffff00000008c6a4 t bpf_prog_29c7ae234d79bd5c_sys_exit [bpf] - Launch any program which accesses file system frequently so can hit the system calls trace flow with eBPF event; - Capture CoreSight trace data with filtering eBPF program: root@debian:~# perf record -e cs_etm/@tmc_etr0/ \ --filter 'filter 0xffff00000008a0d0/0x800' -a sleep 5s - Decode the eBPF program symbol 'bpf_prog_f173133dc38ccf87_sys_enter': root@debian:~# perf script -F,ip,sym Frame deformatter: Found 4 FSYNCS 0 [unknown] ffff00000008a1ac bpf_prog_e470211b846088d5_sys_enter ffff00000008a250 bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a124 bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a14c bpf_prog_e470211b846088d5_sys_enter ffff00000008a13c bpf_prog_e470211b846088d5_sys_enter ffff00000008a14c bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a180 bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a1ac bpf_prog_e470211b846088d5_sys_enter ffff00000008a190 bpf_prog_e470211b846088d5_sys_enter ffff00000008a1ac bpf_prog_e470211b846088d5_sys_enter ffff00000008a250 bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a124 bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a14c bpf_prog_e470211b846088d5_sys_enter 0 [unknown] ffff00000008a180 bpf_prog_e470211b846088d5_sys_enter [...] Cc: Mathieu Poirier <mathieu.poirier(a)linaro.org> Cc: Alexander Shishkin <alexander.shishkin(a)linux.intel.com> Cc: Jiri Olsa <jolsa(a)redhat.com> Cc: Namhyung Kim <namhyung(a)kernel.org> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Suzuki Poulouse <suzuki.poulose(a)arm.com> Cc: Adrian Hunter <adrian.hunter(a)intel.com> Cc: coresight(a)lists.linaro.org Cc: linux-arm-kernel(a)lists.infradead.org Signed-off-by: Leo Yan <leo.yan(a)linaro.org> --- tools/perf/Makefile.config | 22 ++++++++++++++++++++++ tools/perf/util/machine.c | 15 ++++++++++++++- 2 files changed, 36 insertions(+), 1 deletion(-) diff --git a/tools/perf/Makefile.config b/tools/perf/Makefile.config index e4988f49ea79..d7ff839d8b20 100644 --- a/tools/perf/Makefile.config +++ b/tools/perf/Makefile.config @@ -48,9 +48,20 @@ ifeq ($(SRCARCH),x86) NO_PERF_REGS := 0 endif +ARM_PRE_START_SIZE := 0 + ifeq ($(SRCARCH),arm) NO_PERF_REGS := 0 LIBUNWIND_LIBS = -lunwind -lunwind-arm + ifneq ($(wildcard $(srctree)/arch/$(SRCARCH)/kernel/vmlinux.lds),) + # Extract info from lds: + # . = ((0xC0000000)) + 0x00208000; + # ARM_PRE_START_SIZE := 0x00208000 + ARM_PRE_START_SIZE := $(shell egrep ' \. \= ${2}0x[0-9a-fA-F]+${2}' \ + $(srctree)/arch/$(SRCARCH)/kernel/vmlinux.lds | \ + sed -e 's/[(|)|.|=|+|<|;|-]//g' -e 's/ \+/ /g' -e 's/^[ \t]*//' | \ + awk -F' ' '{printf "0x%x", $$2}' 2>/dev/null) + endif endif ifeq ($(SRCARCH),arm64) @@ -58,8 +69,19 @@ ifeq ($(SRCARCH),arm64) NO_SYSCALL_TABLE := 0 CFLAGS += -I$(OUTPUT)arch/arm64/include/generated LIBUNWIND_LIBS = -lunwind -lunwind-aarch64 + ifneq ($(wildcard $(srctree)/arch/$(SRCARCH)/kernel/vmlinux.lds),) + # Extract info from lds: + # . = ((((((((0xffffffffffffffff)) - (((1)) << (48)) + 1) + (0)) + (0x08000000))) + (0x08000000))) + 0x00080000; + # ARM_PRE_START_SIZE := (0x08000000 + 0x08000000 + 0x00080000) = 0x10080000 + ARM_PRE_START_SIZE := $(shell egrep ' \. \= ${8}0x[0-9a-fA-F]+${2}' \ + $(srctree)/arch/$(SRCARCH)/kernel/vmlinux.lds | \ + sed -e 's/[(|)|.|=|+|<|;|-]//g' -e 's/ \+/ /g' -e 's/^[ \t]*//' | \ + awk -F' ' '{printf "0x%x", $$6+$$7+$$8}' 2>/dev/null) + endif endif +CFLAGS += -DARM_PRE_START_SIZE=$(ARM_PRE_START_SIZE) + ifeq ($(SRCARCH),csky) NO_PERF_REGS := 0 endif diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index f6ee7fbad3e4..e993f891bb82 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -2687,13 +2687,26 @@ int machine__get_kernel_start(struct machine *machine) machine->kernel_start = 1ULL << 63; if (map) { err = map__load(map); + if (err) + return err; + /* * On x86_64, PTI entry trampolines are less than the * start of kernel text, but still above 2^63. So leave * kernel_start = 1ULL << 63 for x86_64. */ - if (!err && !machine__is(machine, "x86_64")) + if (!machine__is(machine, "x86_64")) machine->kernel_start = map->start; + + /* + * On arm/arm64, the kernel uses some memory regions which are + * prior to '_stext' symbol; to reflect the complete kernel + * address space, compensate these pre-defined regions for + * kernel start address. + */ + if (!strcmp(perf_env__arch(machine->env), "arm") || + !strcmp(perf_env__arch(machine->env), "arm64")) + machine->kernel_start -= ARM_PRE_START_SIZE; } return err; } -- 2.17.1

6 years, 2 months

[RFC PATCH 0/5] Describe CoreSight topology using sysfs links.

by Mike Leach

The connections between CoreSight sources, links and sinks is not obvious without documentation or access to the device tree / ACPI definitions for the platform. This patchset provides sysfs links to enable the user to follow the trace patch from source to sink. Components in the trace path are updated to have a connections sysfs group, which collates all the links for that component. This supplements the sysfs links created for connections between CTI devices and applies on top of that work from the patchset [1]. Adaptation of an original patchset [2] from Suzuki, reusing 2 patches unchanged with update to 3rd adapt to the CTI v4 RFC set and new code to add default connections group. Tested on DB410c 5.3-rc1, coresight/next branch. [1] https://lists.linaro.org/pipermail/coresight/2019-August/003146.html [2] https://lists.linaro.org/pipermail/coresight/2019-May/002803.html Mike Leach (2): coresight: Move common sysfs functions into separate file. coresight: add default connections group in sysfs to components. Suzuki K Poulose (3): coresight: Pass coresight_device for coresight_release_platform_data coresight: add return value for fixup connections coresight: Expose device connections via sysfs drivers/hwtracing/coresight/Makefile | 3 +- drivers/hwtracing/coresight/coresight-etm3x.c | 1 + drivers/hwtracing/coresight/coresight-etm4x.c | 1 + .../hwtracing/coresight/coresight-funnel.c | 1 + .../hwtracing/coresight/coresight-platform.c | 2 +- drivers/hwtracing/coresight/coresight-priv.h | 10 +- .../coresight/coresight-replicator.c | 1 + drivers/hwtracing/coresight/coresight-stm.c | 1 + drivers/hwtracing/coresight/coresight-sysfs.c | 232 ++++++++++++++++++ drivers/hwtracing/coresight/coresight-tmc.c | 1 + drivers/hwtracing/coresight/coresight.c | 146 ++++------- include/linux/coresight.h | 7 + 12 files changed, 306 insertions(+), 100 deletions(-) create mode 100644 drivers/hwtracing/coresight/coresight-sysfs.c -- 2.20.1

6 years, 2 months

[PATCH 0/8] coresight: etm4x: Fixes and updates for sysfs API

by Mike Leach

Review of ETMV4 sysfs code resulted in a number of minor issues being discovered. Patch set fixes these issues:- 1) Update for ETM v4.4 archtecture. 2) Add missing single shot comparator API. 3) Misc fixes and improvements to sysfs API 4) Updated programmers documentation and reference. Mike Leach (8): coresight: etm4x: Fixes for ETM v4.4 architecture updates. coresight: etm4x: Fix input validation for sysfs. coresight: etm4x: Add missing API to set EL match on address filters coresight: etm4x: Fix issues with start-stop logic. coresight: etm4x: Improve usability of sysfs API. coresight: etm4x: Add view comparator settings API to sysfs. coresight: etm4x: Add missing single-shot control API to sysfs coresight: etm4x: docs: Additional documentation for ETM4x. .../testing/sysfs-bus-coresight-devices-etm4x | 183 ++++--- .../{ => coresight}/coresight-cpu-debug.txt | 0 .../coresight/coresight-etm4x-reference.txt | 459 ++++++++++++++++++ .../trace/{ => coresight}/coresight.txt | 0 MAINTAINERS | 3 +- .../coresight/coresight-etm4x-sysfs.c | 308 +++++++++++- drivers/hwtracing/coresight/coresight-etm4x.c | 32 +- drivers/hwtracing/coresight/coresight-etm4x.h | 18 +- 8 files changed, 902 insertions(+), 101 deletions(-) rename Documentation/trace/{ => coresight}/coresight-cpu-debug.txt (100%) create mode 100644 Documentation/trace/coresight/coresight-etm4x-reference.txt rename Documentation/trace/{ => coresight}/coresight.txt (100%) -- 2.17.1

6 years, 2 months

[PATCH 1/2] perf cs-etm: Support sample flags 'insn' and 'insnlen'

by Leo Yan

The synthetic branch and instruction samples are missed to set instruction related info, thus perf tool fails to display samples with flags '-F,+insn,+insnlen'. CoreSight trace decoder has provided sufficient information to decide the instruction size based on the isa type: A64/A32 instruction are 32-bit size, but one exception is the T32 instruction size, which might be 32-bit or 16-bit. This patch handles for these cases and it reads the instruction values from DSO file; thus can support flags '-F,+insn,+insnlen'. Before: # perf script -F,insn,insnlen,ip,sym 0 [unknown] ilen: 0 ffff97174044 _start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 [...] After: # perf script -F,insn,insnlen,ip,sym 0 [unknown] ilen: 0 ffff97174044 _start ilen: 4 insn: 2f 02 00 94 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 [...] Cc: Mathieu Poirier <mathieu.poirier(a)linaro.org> Cc: Suzuki Poulouse <suzuki.poulose(a)arm.com> Cc: Mike Leach <mike.leach(a)linaro.org> Cc: Robert Walker <robert.walker(a)arm.com> Cc: coresight(a)lists.linaro.org Cc: linux-arm-kernel(a)lists.infradead.org Signed-off-by: Leo Yan <leo.yan(a)linaro.org> --- tools/perf/util/cs-etm.c | 35 ++++++++++++++++++++++++++++++++++- 1 file changed, 34 insertions(+), 1 deletion(-) diff --git a/tools/perf/util/cs-etm.c b/tools/perf/util/cs-etm.c index ed6f7fd5b90b..b3a5daaf1a8f 100644 --- a/tools/perf/util/cs-etm.c +++ b/tools/perf/util/cs-etm.c @@ -1076,6 +1076,35 @@ bool cs_etm__etmq_is_timeless(struct cs_etm_queue *etmq) return !!etmq->etm->timeless_decoding; } +static void cs_etm__copy_insn(struct cs_etm_queue *etmq, + u64 trace_chan_id, + const struct cs_etm_packet *packet, + struct perf_sample *sample) +{ + /* + * It's pointless to read instructions for the CS_ETM_DISCONTINUITY + * packet, so directly bail out with 'insn_len' = 0. + */ + if (packet->sample_type == CS_ETM_DISCONTINUITY) { + sample->insn_len = 0; + return; + } + + /* + * T32 instruction size might be 32-bit or 16-bit, decide by calling + * cs_etm__t32_instr_size(). + */ + if (packet->isa == CS_ETM_ISA_T32) + sample->insn_len = cs_etm__t32_instr_size(etmq, trace_chan_id, + sample->ip); + /* Otherwise, A64 and A32 instruction size are always 32-bit. */ + else + sample->insn_len = 4; + + cs_etm__mem_access(etmq, trace_chan_id, sample->ip, + sample->insn_len, (void *)sample->insn); +} + static int cs_etm__synth_instruction_sample(struct cs_etm_queue *etmq, struct cs_etm_traceid_queue *tidq, u64 addr, u64 period) @@ -1097,9 +1126,10 @@ static int cs_etm__synth_instruction_sample(struct cs_etm_queue *etmq, sample.period = period; sample.cpu = tidq->packet->cpu; sample.flags = tidq->prev_packet->flags; - sample.insn_len = 1; sample.cpumode = event->sample.header.misc; + cs_etm__copy_insn(etmq, tidq->trace_chan_id, tidq->packet, &sample); + if (etm->synth_opts.last_branch) { cs_etm__copy_last_branch_rb(etmq, tidq); sample.branch_stack = tidq->last_branch; @@ -1159,6 +1189,9 @@ static int cs_etm__synth_branch_sample(struct cs_etm_queue *etmq, sample.flags = tidq->prev_packet->flags; sample.cpumode = event->sample.header.misc; + cs_etm__copy_insn(etmq, tidq->trace_chan_id, tidq->prev_packet, + &sample); + /* * perf report cannot handle events without a branch stack */ -- 2.17.1

6 years, 2 months

DHL EXPRESS DELIVERY NOTIFICATION

by DHL International GmbH

6 years, 2 months

[PATCH 11/17] perf cs-etm: Support sample flags 'insn' and 'insnlen'

by Arnaldo Carvalho de Melo

From: Leo Yan <leo.yan(a)linaro.org> The synthetic branch and instruction samples are missed to set instruction related info, thus the perf tool fails to display samples with flags '-F,+insn,+insnlen'. The CoreSight trace decoder provides sufficient information to decide the instruction size based on the ISA type: A64/A32 instructions are 32-bit size, but one exception is the T32 instruction size, which might be 32-bit or 16-bit. This patch handles these cases and it reads the instruction values from DSO file; thus can support the flags '-F,+insn,+insnlen'. Before: # perf script -F,insn,insnlen,ip,sym 0 [unknown] ilen: 0 ffff97174044 _start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 ffff97174938 _dl_start ilen: 0 [...] After: # perf script -F,insn,insnlen,ip,sym 0 [unknown] ilen: 0 ffff97174044 _start ilen: 4 insn: 2f 02 00 94 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 ffff97174938 _dl_start ilen: 4 insn: c1 ff ff 54 [...] Signed-off-by: Leo Yan <leo.yan(a)linaro.org> Reviewed-by: Mathieu Poirier <mathieu.poirier(a)linaro.org> Tested-by: Mathieu Poirier <mathieu.poirier(a)linaro.org> Cc: Alexander Shishkin <alexander.shishkin(a)linux.intel.com> Cc: Jiri Olsa <jolsa(a)redhat.com> Cc: Mike Leach <mike.leach(a)linaro.org> Cc: Namhyung Kim <namhyung(a)kernel.org> Cc: Robert Walker <robert.walker(a)arm.com> Cc: Suzuki Poulouse <suzuki.poulose(a)arm.com> Cc: coresight(a)lists.linaro.org Cc: linux-arm-kernel(a)lists.infradead.org Link: http://lkml.kernel.org/r/20190815082854.18191-1-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo <acme(a)redhat.com> --- tools/perf/util/cs-etm.c | 35 ++++++++++++++++++++++++++++++++++- 1 file changed, 34 insertions(+), 1 deletion(-) diff --git a/tools/perf/util/cs-etm.c b/tools/perf/util/cs-etm.c index ed6f7fd5b90b..b3a5daaf1a8f 100644 --- a/tools/perf/util/cs-etm.c +++ b/tools/perf/util/cs-etm.c @@ -1076,6 +1076,35 @@ bool cs_etm__etmq_is_timeless(struct cs_etm_queue *etmq) return !!etmq->etm->timeless_decoding; } +static void cs_etm__copy_insn(struct cs_etm_queue *etmq, + u64 trace_chan_id, + const struct cs_etm_packet *packet, + struct perf_sample *sample) +{ + /* + * It's pointless to read instructions for the CS_ETM_DISCONTINUITY + * packet, so directly bail out with 'insn_len' = 0. + */ + if (packet->sample_type == CS_ETM_DISCONTINUITY) { + sample->insn_len = 0; + return; + } + + /* + * T32 instruction size might be 32-bit or 16-bit, decide by calling + * cs_etm__t32_instr_size(). + */ + if (packet->isa == CS_ETM_ISA_T32) + sample->insn_len = cs_etm__t32_instr_size(etmq, trace_chan_id, + sample->ip); + /* Otherwise, A64 and A32 instruction size are always 32-bit. */ + else + sample->insn_len = 4; + + cs_etm__mem_access(etmq, trace_chan_id, sample->ip, + sample->insn_len, (void *)sample->insn); +} + static int cs_etm__synth_instruction_sample(struct cs_etm_queue *etmq, struct cs_etm_traceid_queue *tidq, u64 addr, u64 period) @@ -1097,9 +1126,10 @@ static int cs_etm__synth_instruction_sample(struct cs_etm_queue *etmq, sample.period = period; sample.cpu = tidq->packet->cpu; sample.flags = tidq->prev_packet->flags; - sample.insn_len = 1; sample.cpumode = event->sample.header.misc; + cs_etm__copy_insn(etmq, tidq->trace_chan_id, tidq->packet, &sample); + if (etm->synth_opts.last_branch) { cs_etm__copy_last_branch_rb(etmq, tidq); sample.branch_stack = tidq->last_branch; @@ -1159,6 +1189,9 @@ static int cs_etm__synth_branch_sample(struct cs_etm_queue *etmq, sample.flags = tidq->prev_packet->flags; sample.cpumode = event->sample.header.misc; + cs_etm__copy_insn(etmq, tidq->trace_chan_id, tidq->prev_packet, + &sample); + /* * perf report cannot handle events without a branch stack */ -- 2.21.0

6 years, 2 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

CoreSight