Perf-Report Show Value of CPU Register

perf-report show value of CPU register

Try perf script data dumping command with the iregs field: perf script -F ip,sym,iregs. All fields -F are documented with source code of tools/perf/builtin-script.c - struct output_option .. all_output_options, and iregs is still here (also OPT_CALLBACK('F', "fields" ... in same file). There is also perf-script.txt documentation - https://github.com/torvalds/linux/blob/master/tools/perf/Documentation/perf-script.txt#L115

-F::
--fields::
    Comma separated list of fields to print. Options are:
    comm, tid, pid, time, cpu, event, trace, ip, sym, dso, addr, symoff,
    srcline, period, iregs, brstack, brstacksym, flags, bpf-output, brstackinsn,
    callindent, insn, insnlen. Field list can be prepended with the type, trace, sw or hw,
    to indicate to which event type the field list applies.

---

There were commits into linux kernel git which mention flag --intr-regs. Start from option implementation:

https://github.com/torvalds/linux/search?utf8=%E2%9C%93&q=intr-regs&type=

tools/perf/builtin-record.c
OPT_CALLBACK_OPTARG('I', "intr-regs", &record.opts.sample_intr_regs, NULL, "any register",

Then search for 'sample_intr_regs' in commits: https://github.com/torvalds/linux/search?q=sample_intr_regs&type=Commits

Several commits were about kernel part and perf_attr debug print. But this has example of intr-regs printing (Sep 1, 2015)
https://github.com/torvalds/linux/commit/532026612455a4a6fd27c1b2e7111263f63218a2

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
- Add ability to specify to select which registers to record,
to reduce the size of perf.data files, and also allow printing
the registers in 'perf script': (Stephane Eranian)

  # perf record --intr-regs=AX,SP usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.016 MB perf.data (8 samples) ]
  # perf script -F ip,sym,iregs | tail -5
   ffffffff8105f42a native_write_msr_safe   AX:0xf    SP:0xffff8802629c3c00
   ffffffff8105f42a native_write_msr_safe   AX:0xf    SP:0xffff8802629c3c00
   ffffffff81761ac0 _raw_spin_lock   AX:0xffff8801bfcf8020    SP:0xffff8802629c3ce8
   ffffffff81202bf8 __vma_adjust_trans_huge   AX:0x7ffc75200000    SP:0xffff8802629c3b30
   ffffffff8122b089 dput   AX:0x101    SP:0xffff8802629c3c78
  #

Recompile Linux Kernel not to use specific CPU register

You should not try to recompile kernel when you just want to sample some register with perf. As I understand, kernel has its own set of registers and will not overwrite user R11. syscall interface uses some fixed registers which can't be changed (can you try different reg?) and there are often glibc gateways to syscall which may use some additional registers (they are not in kernel, they are user-space code; often generated or written in assembler). You may try using gdb to monitor the register to change to find who did it. It can do this (hmm, one more link to the same user on SO): gdb: breakpoint when register will have value 0xffaa like gdb ./program then gdb commands start; watch $r11; continue; where.

Two weeks age there was question perf-report show value of CPU register about register value sampling with perf:

I follow this document and using perf record with --intr-regs=ax,bx,r15, trying to log additional CPU register information with PEBS record.

While there was x86 & PEBS, ARM may have --intr-regs implemented too. Check output of perf record --intr-regs=\? (man perf-record: "To list the available registers use --intr-regs=\?") to find support status and register names.

To print registers, use perf script -F ip,sym,iregs command. There was example in some linux commits:

  # perf record --intr-regs=AX,SP usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.016 MB perf.data (8 samples) ]
  # perf script -F ip,sym,iregs | tail -5
   ffffffff8105f42a native_write_msr_safe   AX:0xf    SP:0xffff8802629c3c00
   ffffffff8105f42a native_write_msr_safe   AX:0xf    SP:0xffff8802629c3c00
   ffffffff81761ac0 _raw_spin_lock   AX:0xffff8801bfcf8020    SP:0xffff8802629c3ce8
   ffffffff81202bf8 __vma_adjust_trans_huge   AX:0x7ffc75200000    SP:0xffff8802629c3b30
   ffffffff8122b089 dput   AX:0x101    SP:0xffff8802629c3c78
  #

Why is moving data between two CPU registers so slow that it costs 30% of total time?

That's not moving between rax and rdx.

That's indexing into an array pointed to by rax by rdx and putting the result in rax. Probable L1 cache miss.

Meaning of time measures given by Linux perf profiler

I found this on the perf wiki:

Interrupt-based sampling introduces skids on modern processors. That
means that the instruction pointer stored in each sample designates
the place where the program was interrupted to process the PMU
interrupt, not the place where the counter actually overflows, i.e.,
where it was at the end of the sampling period. In some case, the
distance between those two points may be several dozen instructions or
more if there were taken branches. When the program cannot make
forward progress, those two locations are indeed identical. For this
reason, care must be taken when interpreting profiles.

That may be the explanation. Unfortunately, the wiki does not say how to figure out if this is indeed the problem or how to correct for this issue.

---

Related Topics