Commits

Show all
Author Commit Message Labels Comments Date
Armin Rigo
Tweaks, and implementation of a wait-free, best-effort algorithm to handle multiple threads (some possibly in signals) that all try to write chunks of data to the same file
Branches
vmprof-review
Richard Plangger
setting the orig label arguments in the test case
Branches
vecopt
Richard Plangger
failargs can be none
Branches
vecopt
Remi Meier
produce the perf-PID.map directly from pypy Produce the jitted code to trace name mapping for perf directly in PyPy. Should probably become a profagent in the future.
Branches
stmgc-c8-gcc
Richard Plangger
could end in a finish operation, crashed the optimizer
Branches
vecopt
Ronan Lamy
Clean up accumulate loops
Branches
ufunc-reduce
Ronan Lamy
Clean up shape handling in ufunc.accumulate()
Branches
ufunc-reduce
Ronan Lamy
Remove wrong docstring
Branches
ufunc-reduce
Armin Rigo
Forgot to call the remove_*() functions in disable()
Branches
vmprof-review
mattip
test, fix for issue #2105
Armin Rigo
Enable(), disable()
Branches
vmprof-review
Armin Rigo
Tweaks tweaks until the tests pass
Branches
vmprof-review
Armin Rigo
Add some files from pypy/module/_vmprof/src
Branches
vmprof-review
Armin Rigo
Test and fixes for enable()
Branches
vmprof-review
Armin Rigo
in-progress
Branches
vmprof-review
Armin Rigo
Skip running the tests based on platform
Branches
vmprof-review
Armin Rigo
Pass the translation tests for the generated trampolines
Branches
vmprof-review
Armin Rigo
in-progress
Branches
vmprof-review
Armin Rigo
hg merge default
Branches
vmprof-review
Armin Rigo
tweaks
Branches
vmprof-review
Armin Rigo
Add another sanity test: rarithmetic.maxint == sys.maxint for now
Armin Rigo
In case this mmap() fails with a different error code than ENOMEM, complain explicitly.
Armin Rigo
Workaround for issue #2102. Can't write a small test for this case, and I've no clue if in this case it means that invalid codemap was produced earlier.
Carl Friedrich Bolz
a bit just because, use sets instead of dicts with None/True values
Armin Rigo
This line looks suspiciously like a leftover from testing
Armin Rigo
Issue #2095: test and fix
Richard Plangger
once again I did not copy the descr...
Branches
vecopt
Richard Plangger
that was a nasty problem. entering the vecopt trace through the preamble only worked for non accum/expanded traces, otherwise the arguments would not match the loop has now an original label, where invariant operations follow leading to an label that can carry expanded values
Branches
vecopt
Remi Meier
make script for generating perf-maps "good-enough" Remove some features to simplify the script a lot
Branches
stmgc-c8-gcc
mattip
translation fixes for win32
Branches
py3k
  1. Prev
  2. Next