On Tue, Jul 05, 2005 at 08:52:25AM +0200, Jacek Piskozub wrote: >I'll risk wasting some more precious time (ypurs or mine) adding a Perl >stackdump I found in the Mozilla directory where the build failed: > >===begin stackdump==== >Exception: STATUS_ACCESS_VIOLATION at eip=610AA95A >eax=2F632F65 ebx=08000000 ecx=871C9010 edx=871C9000 esi=2F632F65 >edi=871C26C8 >ebp=0074E998 esp=0074E970 program=C:\CYGWIN\BIN\PERL.EXE, pid 975019, >thread main >cs=018F ds=0197 es=0197 fs=51EF gs=0000 ss=0197 >Stack trace: >Frame Function Args >0074E998 610AA95A (00000000, 10140168, 0074E9C8, 6109CD52) >0074E9C8 6104EC66 (871C3640, 102B2EC8, 0074E9F8, 6109CD52) >0074E9F8 610844FF (101401C8, 102B5280, 0074EA28, 101255F0) >0074EA28 10032A1F (101401C8, 102B5AC0, 0074EA58, 6109CD52) >0074EA58 10032A1F (101401C8, 102B5AE0, 0074EA88, 101255F0) >0074EA88 10032A1F (101401C8, 102B5B00, 0074EAB8, 6109CD52) >0074EAB8 10032A1F (101401C8, 102B5B20, 10169050, 101255F0) >0074EAE8 10032A1F (101401C8, 102B5B40, 0074EB18, 6109CD52) >0074EB18 10032A1F (101401C8, 102B5B60, 0074EB48, 6109CD52) >0074EB48 10032A1F (101401C8, 102B5B80, 0074EB78, 101255F0) >0074EB78 10032A1F (101401C8, 102B4FE0, 0074EBA8, 6109CD52) >0074EBA8 10032A1F (101401C8, 102B5BE0, 0074EBD8, 6109CD52) >0074EBD8 10032A1F (101401C8, 102B5C00, 00000000, FFFFFFFF) >0074EC08 10032A1F (101401C8, 10143CC0, 102DA2A4, 101410B8) >0074ED58 10008053 (101401C8, 00401220, 00000012, 61157CB8) >0074ED98 00401146 (00000012, 61157CB8, 10140090, 00000000) >End of stack trace (more stack frames may be present) >====end stackdump=====
This is dying in malloc. That's usually a symptom of a "double free". This is, AFAICT, unrelated to Volker's problem since there is no way that my fix from last night would affect a malloc problem. The malloc implementation in perl was updated to the latest version of Doug Lea's malloc so there is a possibility that this is a malloc bug. However, given that the perl on my linux system routinely trips a "double free" error from my glibc library, my money would be on perl. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/