632402 - Valgrind failures in acceptance suite

Status: RESOLVED WONTFIX

(Tamarin Graveyard :: Garbage Collection (mmGC), defect, P4)

Description

[Brent Baker]

There are a couple of testcases that are triggering valgrind errors when running 32bit Release-Debugger builds (64bit builds do not trigger these errors)

compiled: ../configure.py --enable-shell --target=i686-darwin --mac-sdk=104u --enable-debugger --enable-valgrind

./runtests.py --valgrind as3/Definitions/FunctionAccessors/InfRec.as spidermonkey/js1_5/Exceptions/regress-121658.as spidermonkey/js1_5/Regress/regress-234389.as

Comment 1

[Brent Baker]

Attachment: valgrind log (as3/Definitions/FunctionAccessors/InfRec.as)

./runtests.py --valgrind as3/Definitions/FunctionAccessors/InfRec.as 

==8840== Invalid write of size 4
==8840==    at 0x593F: VMPI_cleanStack(unsigned long)

==8840== Invalid read of size 4
==8840==    at 0x2B5A46: memset (mc_replace_strmem.c:626)
==8840==    by 0x5946: VMPI_cleanStack(unsigned long)

Comment 2

[Brent Baker]

(In reply to comment #1)
> Created attachment 510622 [details]
> valgrind log (as3/Definitions/FunctionAccessors/InfRec.as)

Other 2 testcases produce the same read/write errors in VMPI_cleanStack()

Comment 3

[Tommy Reilly]

weird, doesn't look like our problem but should be investigated further, just add these tests to the valgrind exclusions for now

Comment 4

[Brent Baker]

I am able to reproduce this failure on linux 32bit.

mac and linux 64bit builds do not have this failure.

Comment 5

[Brent Baker]

Attachment: Mark expected failures

Comment 6

[Tamarin Bot]

changeset: 5906:bdc26f8b613c
user:      Brent Baker <[e-mail]>
summary:   Bug 632402: mark expected failure of testcases when running 32bit shell through valgrind (r=cpeyer)

http://hg.mozilla.org/tamarin-redux/rev/bdc26f8b613c

Comment 7

[Sylvestre Ledru [:Sylvestre]]

Tamarin is a dead project now. Mass WONTFIX.

Comment 8

[Sylvestre Ledru [:Sylvestre]]

Tamarin isn't maintained anymore. WONTFIX remaining bugs.