New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
map leaks memory #9294
Comments
From elmex@ta-sa.orgThis is a bug report for perl from elmex@ta-sa.org. This perl code leaks memory on all perl 5.10 I had access to: while (1) { map 1, 1 } Flags: Site configuration information for perl 5.10.0: Configured by root at Wed Dec 19 23:01:57 CET 2007. Summary of my perl5 (revision 5 version 10 subversion 0) configuration: Locally applied patches: @INC for perl 5.10.0: Environment for perl 5.10.0: |
From @shlomifHi! On Friday 18 April 2008, Robin Redeker wrote:
I can't see how I can comment on the bug, but I'll note that I can confirm it Also: {{{{{{{{{{{{{{{{{ I should note that I've been on the IRC when Robin reported this bug there. Regards, Shlomi Fish
-- Shlomi Fish shlomif@iglu.org.il I'm not an actor - I just play one on T.V. |
The RT System itself - Status changed from 'new' to 'open' |
From @nwc10On Sat, Apr 19, 2008 at 09:44:02AM +0300, Shlomi Fish wrote:
Which says: ==13629== 1,370,724 bytes in 649 blocks are still reachable in loss record 15 of 15 which appears to be bogus, because if I stick a breakpoint, here are all (gdb) r Breakpoint 1, S_more_bodies (sv_type=SVt_PV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVNV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVAV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVHV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVIV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVGV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVMG) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVCV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVCV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVGV) at sv.c:1047 Breakpoint 1, S_more_bodies (sv_type=SVt_PVIO) at sv.c:1047 That isn't 649 blocks.
^^^^^ doesn't leak under ithreads Nicholas Clark |
From @nwc10On Sat, Apr 19, 2008 at 10:34:53AM +0100, Nicholas Clark wrote:
sticking a breakpoint on all the malloc functions seems interesting Breakpoint 2, Perl_safesysmalloc (size=4080) at util.c:89 Breakpoint 2, Perl_safesysmalloc (size=4080) at util.c:89 And, on a macine where hardware watchpoints actually (bloody) work: Breakpoint 1, Perl_pp_grepstart () at pp_ctl.c:951 Old value = (struct sv *) 0x0 Old value = (struct sv *) 0x8524e0 Old value = (struct sv *) 0x8655e8 Breakpoint 1, Perl_pp_grepstart () at pp_ctl.c:951 Old value = (struct sv *) 0x0 Old value = (struct sv *) 0x852150 Old value = (struct sv *) 0x8655e8 we're leaking the SV that is assigned to DEFSV in this part of pp_grepstart: SAVETMPS; src = PL_stack_base[*PL_markstack_ptr]; I'm not actually sure whether we are leaking it, or merely "leaking" it
Does leak under ithreads, but it seems that it clears up before valgrind gets Nicholas Clark |
From p5p@spam.wizbit.beOn Fri, 18 Apr 2008 14:18:52 +0200, Robin Redeker
Could this be/Is this related to: [perl #48004] unacceptable memory consumption Kind regards, Bram |
From szbalint@inf.elte.huI went on a historical field-trip and tried to check what changed http://xrl.us/ooeip Reverting to the 5.8.8 behaviour by defining PERL_CREATE_GVSV makes this This should narrow the scope of tracking down this leak, however I'm not |
From @mhxFixed by the following change: http://public.activestate.com/cgi-bin/perlbrowse/p/34776 34776 on 2008/11/08 by mhx@mhx-r2d2 Assigning to DEFSV leaks if PL_defgv's gp_sv isn't set. [perl #48004] is a completely different problem. |
@mhx - Status changed from 'open' to 'resolved' |
Migrated from rt.perl.org#53038 (status was 'resolved')
Searchable as RT53038$
The text was updated successfully, but these errors were encountered: