Adds support for cpu profiler logging to the interpreter. Modifies the
the API to be passed AbstractCode objects instead of Code objects, and
adds extra functions to AbstractCode which is required by log.cc and
cpu-profiler.cc.
The main change in sampler.cc is to determine if a stack frame is an
interpreter stack frame, and if so, use the bytecode address as the pc
for that frame. This allows sampling of bytecode functions. This
requires adding support to SafeStackIterator to determine if a frame is
interpreted, which we do by checking the PC against pre-stored addresses
for the start and end of interpreter entry builtins.
Also removes CodeDeleteEvents which are dead code and haven't
been reported for some time.
Still to do is tracking source positions which will be done in a
followup CL.
BUG=v8:4766
LOG=N
Review URL: https://codereview.chromium.org/1728593002
Cr-Commit-Position: refs/heads/master@{#34321}
This simplifies correlating offsets with the output from
--print-opt-code, which outputs offsets in decimal.
We keep the hex output since branch instructions in the perf dump use
hex labels. We just include the decimal offset along with the hex offset.
BUG=
Review URL: https://codereview.chromium.org/1612403002
Cr-Commit-Position: refs/heads/master@{#33455}
Show tick count, besides the percentage spent on an
instruction. Aids perf investigations where we deal with
stalls, for example. Percentage-wise, the execution appears
distributed similarly, but the regression becomes more
apparent in the tick counts.
Review URL: https://codereview.chromium.org/1607323003
Cr-Commit-Position: refs/heads/master@{#33452}
Newer perf.data contains both MMAP and MMAP2 record type,
but MMAP2 record type is not supported in previous ll_prof,
MMAP2 record information will be lost.
BUG=v8:4569
LOG=n
Review URL: https://codereview.chromium.org/1469153004
Cr-Commit-Position: refs/heads/master@{#32319}
Switched to using binary low-level log instead of the textual log used
by the ticks processor. The binary log contains code-related events,
code object names, and their bodies. When writing to the log we ask
glibc to use a larger buffer. To avoid complex processing of the
snapshot log (which is still textual) the serializer emits final
snapshot position to code name mappings that can be quickly be read
without replaying the snapshot log. (This might be useful for the
ticks processor.)
Review URL: http://codereview.chromium.org/6904127
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@7729 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
The main issue was due to multiple recompilations of functions. Now
code objects are grouped by function using SFI object address.
JSFunction objects are no longer tracked, instead we track SFI object
moves. To pick a correct code version, we now sample return addresses
instead of JSFunction addresses.
tools/{linux|mac|windows}-tickprocessor scripts differentiate
between code optimization states for the same function
(using * and ~ prefixes introduced earlier).
DevTools CPU profiler treats all variants of function code as
a single function.
ll_prof treats each optimized variant as a separate entry, because
it can disassemble each one of them.
tickprocessor.py not updated -- it is deprecated and will be removed.
BUG=v8/1087,b/3178160
TEST=all existing tests pass, including Chromium layout tests
Review URL: http://codereview.chromium.org/6551011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6902 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Analyses full minidump (.dmp) files.
Shows the processor state at the point of exception including the
stack of the active thread and the referenced objects in the V8
heap. Code objects are disassembled and the addresses linked from the
stack (pushed return addresses) are marked with "=>".
Review URL: http://codereview.chromium.org/6312058
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@6896 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Since 2.6.31 perf_events interface has been available in the
kernel. There's a nice tool called "perf" (linux-2.6/tools/perf) that
uses this interface and provides capabilities similar to oprofile. The
simplest form of its usage is just dumping the raw log (trace) of
events generated by the kernel. In this patch I'm adding a script
(tools/ll_prof.py) to build profiles based on perf trace and our code
log. All the heavy-lifting is done by perf. Compared to oprofile agent
this approach does not require recompilation and supports code moving
garbage collections.
Expected usage is documented in the ll_prof's help. Basically one
should run V8 under perf passing --ll-prof flag and then the produced
logs can be analyzed by tools/ll_prof.py.
The new --ll-prof flag enables logging of generated code object
locations and names (like --log-code), and also of their bodies, which
can be later disassembled and annotated by the script.
Review URL: http://codereview.chromium.org/3831002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5663 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
