#282 Msys 1.0.9 make crashes

closed-fixed
Earnie Boyd
MSYS (75)
2014-12-29
2003-02-12
Johnny Willemsen
No

Hi,

Because of bug 683944 I upgraded to the 1.0.9 pre
release. With this version I have a new crash.

Build log is available at
http://www.cs.wustl.edu/~bugzilla/auto_compile_logs/re
medy.nl_Win2K_MingW/log.txt you can see also there
the Msys version

The stack dump
Exception: STATUS_ACCESS_VIOLATION at
eip=71098031
eax=00000808 ebx=00000004 ecx=00000C00
edx=00000808 esi=0A34FF60 edi=0A34F360
ebp=0022D354 esp=0022D32C program=C:\MSYS\1.0
\BIN\make.exe
cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023
Stack trace:
Frame Function Args
0022D354 71098031 (710A4020, 0A34FE40,
0A08FB08, 0A333C18)
0022D374 710308B5 (0A34FE40, 7109BE50,
0022D27C, 0A333CF0)
0022D394 71030728 (0A34FE40, 0A333BA8,
0022D434, 7109B8EA)
0022D434 7109B925 (0A34FE40, 0A34FE40,
0000001C, 7106174E)
0022E804 71061B22 (00000000, 0022E8B4,
0A322EF0, 0A323448)
0022E844 71063537 (00000000, 00000003,
0022E8B4, 0A322EF0)
0022E874 71010AC9 (0022E8B4, 0A322EF0,
0022ECB4, 7108CFEB)
0022ECB4 7108CFF8 (0A323330, 0A322EF0,
0022ECE4, 0040CB77)
0022ECE4 0040CB89 (0A322EF0, 0A323448,
00425440, C0996B18)
0022ED14 0040D7C0 (00000000, 00000001,
0A322EF0, 0A323448)
0022ED74 0040C37A (0A08F240, 710972A7,
0022EDB4, 00404647)
0022EDB4 0040C51E (0A08F240, 0A0D2F00,
0022EE04, 0040C69F)
0022EE04 0040C9A4 (0A0D2F00, 00000000,
00000000, 0041961C)
0022EE34 00402BE1 (0A0D2F00, 00000000,
00000001, 00000000)
0022EE64 00419367 (0A0D2F00, 00000001,
0000560C, 0017034D)
0022EEC4 00418993 (0A0D2F00, 00000009,
00000000, 00000000)
End of stack trace (more stack frames may be
present) 0 [main] make 1424 sig_send: wait for
sig_complete event failed, signal -2, rc 258, Win32 error
0
60026324 [main] make 1424 sig_send: wait for
sig_complete event failed, signal -2, rc 258, Win32 error
0
g++ -W -Wall -Wpointer-arith -mcpu=pentiumpro -
mthreads -W -Wall -Wpointer-arith -O3 -g -fno-
exceptions -fcheck-new -
ID:/ACE/MingW/ACE_wrappers -
ID:/ACE/MingW/ACE_wrappers/TAO -
DTAO_PORTABLESERVER_BUILD_DLL -c -
o .shobj/Servant_Base.o Servant_Base.cpp
120072632 [main] make 1424 sig_send: wait for
sig_complete event failed, signal -2, rc 258, Win32 error
0
180098946 [main] make 1424 sig_send: wait for
sig_complete event failed, signal -2, rc 258, Win32 error
0

Discussion

1 2 > >> (Page 1 of 2)
  • Earnie Boyd
    Earnie Boyd
    2003-02-12

    • status: open --> pending
     
    • status: pending --> open
     
  • Logged In: YES
    user_id=585332

    Snapshot is installed on the system and the build will run
    this night.

     
  • Logged In: YES
    user_id=585332

    Snapshot is installed on the system and the build will run
    this night.

     
    • status: open --> pending
     
  • Logged In: YES
    user_id=585332

    The build has run but we have a stackdump again. The
    stackdump is
    Exception: STATUS_ACCESS_VIOLATION at eip=710980B1
    eax=00000808 ebx=00000004 ecx=00000C00 edx=00000808
    esi=0A34FF60 edi=0A34F360
    ebp=0022D354 esp=0022D32C program=C:\MSYS\1.0
    \BIN\make.exe
    cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023
    Stack trace:
    Frame Function Args
    0022D354 710980B1 (710A4020, 0A34FE40, 0A0530D8,
    0A333C18)
    0022D374 710308B5 (0A34FE40, 7109BED0, 0022D27C,
    0A333CF0)
    0022D394 71030728 (0A34FE40, 0A333BA8, 0022D434,
    7109B96A)
    0022D434 7109B9A5 (0A34FE40, 0A34FE40, 0000001C,
    710617BE)
    0022E804 71061B92 (00000000, 0022E8B4, 0A322EF0,
    0A323448)
    0022E844 710635B7 (00000000, 00000003, 0022E8B4,
    0A322EF0)
    0022E874 71010AC9 (0022E8B4, 0A322EF0, 0022ECB4,
    7108D06B)
    0022ECB4 7108D078 (0A323330, 0A322EF0, 0022ECE4,
    0040CB77)
    0022ECE4 0040CB89 (0A322EF0, 0A323448, 00425440,
    C0996B18)
    0022ED14 0040D7C0 (00000000, 00000001, 0A322EF0,
    0A323448)
    0022ED74 0040C37A (0A052F38, 71097327, 0022EDB4,
    00404647)
    0022EDB4 0040C51E (0A052F38, 0A0D2F00, 0022EE04,
    0040C69F)
    0022EE04 0040C9A4 (0A0D2F00, 00000000, 00000000,
    0041961C)
    0022EE34 00402BE1 (0A0D2F00, 00000000, 00000001,
    00000000)
    0022EE64 00419367 (0A0D2F00, 00000001, 0000560C,
    002B3297)
    0022EEC4 00418993 (0A0D2F00, 00000009, 00000000,
    00000000)
    End of stack trace (more stack frames may be present)
    62925292 [main] make 1596 sig_send: wait for sig_complete
    event failed, signal -2, rc 258, Win32 error 0
    122951601 [main] make 1596 sig_send: wait for
    sig_complete event failed, signal -2, rc 258, Win32 error 0
    182987913 [main] make 1596 sig_send: wait for
    sig_complete event failed, signal -2, rc 258, Win32 error 0
    243014235 [main] make 1596 sig_send: wait for
    sig_complete event failed, signal -2, rc 258, Win32 error 0

    Also available at
    http://www.cs.wustl.edu/~bugzilla/auto_compile_logs/remedy.
    nl_Win2K_MingW/make.exe.stackdump

    The complete build log is available at
    http://www.cs.wustl.edu/~bugzilla/auto_compile_logs/remedy.
    nl_Win2K_MingW/log.txt

     
    • status: pending --> open
     
  • Earnie Boyd
    Earnie Boyd
    2003-02-18

    • status: open --> pending
     
  • Earnie Boyd
    Earnie Boyd
    2003-02-20

    Logged In: YES
    user_id=15438

    I may have caused other issues, but this one isn't one any
    longer.

    Earnie.

     
  • Earnie Boyd
    Earnie Boyd
    2003-02-20

    • status: pending --> closed-fixed
     
  • Logged In: YES
    user_id=585332

    FYI, the make didn't crash last night in our build

     
  • Logged In: YES
    user_id=585332

    Opened again. With the release of 17 feb and 7 march I get
    sometimes stack dumps again in the daily build. They where
    gone some weeks but now the make crashes daily again.

     
    • status: closed-fixed --> open-fixed
     
  • Logged In: YES
    user_id=585332

    Installed release MSYS-1.0.9-2003.03.07-1.exe and this also
    gives stackdumps during the build in make

     
  • Logged In: YES
    user_id=585332

    I have also tested 14-04-2003 and 15-04-2003 test releases
    of MSYS, but I then get the following errors. Looks
    something is really wrong with that release. When I type awk
    on the commandprompt it just works

    Running: make -k reverseclean
    make: awk: Command not found
    make: awk: Command not found
    make: awk: Command not found
    make: g++: Command not found
    make: g++: Command not found
    /bin/sh.exe: egrep: command not found
    D:/ACE/MingW/ACE_wrappers/bin/reverse_clean: exec: perl:
    not found
    make: *** [reverseclean] Error 127

    make

     
  • Logged In: YES
    user_id=585332

    The release of msys of 17-4-2003 now doesn't give these
    strange errors, but still gives a stackdump during compilation
    of ace/tao.

     
  • Logged In: YES
    user_id=585332

    We have investigated this further. The problem appears when
    linking large shared libraries (up to 70Mb in debug mode/6Mb
    release) from within a perl script. The perl script redirects
    stderr en stdout to a file and runs make. When running make
    directly from the command line it just works fine. We tested
    with multiple perl versions and with all versions the make
    gives the same problem.

     
  • Logged In: YES
    user_id=585332

    We have had one succesfull build but now it is failing again
    with a stackdump in make. I have also got a report from
    someone else that he got a make stackdump with the 2003-
    06-05 msys snapshot, so there is still somewhere a problem.
    This week we will investigate it further and add the
    stackdump info here.

     
  • Logged In: YES
    user_id=585332

    hereby some stackdumps I received from someone else that
    is using ACE/TAO with msys snapshot 2003-06-05

    <STACKDUMP>

    g++ -W -Wall -Wpointer-arith -mcpu=pentiumpro -mthreads -
    W -Wall -Wpointer-a
    rith -O3 -g -Wno-uninitialized -I/d/ACE_wrappers/ -
    DACE_HAS_EXCEPTIONS -
    DACE_RMCAST_BUILD_DLL -c -
    o .shobj/RMCast_Resend_Worker.o
    RMCast_Resend_Worker.cpp
    0 [main] make 3012 open_stackdumpfile: Dumping stack
    trace to
    make.exe.stackdump
    make[2]: *** [RMCast.subdir] Segmentation fault (core
    dumped)
    ..................................................
    Exception: STATUS_ACCESS_VIOLATION at eip=71098C0D
    eax=00000000 ebx=00000000 ecx=00000148 edx=00000000
    esi=0A103E58
    edi=0A103D10
    ebp=0022ED14 esp=0022ECEC
    program=C:\util\msys\bin\make.exe
    cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023
    Stack trace:
    Frame Function Args
    0022ED14 71098C0D (710A5020, 0A103E28, 00000001,
    00000000)
    0022ED34 7103110D (0A103E28, 00080000, 00000000,
    0022ED74)
    0022ED74 0040BEEE (0A039200, 0022EDC0, 0022EDC4,
    0040BCEC)
    0022EDC4 0040BD20 (00000001, 00000000, 0022EE14,
    0040C69F)
    0022EE14 0040C9CF (0A077DA8, 00000000, 00000000,
    0041961C)
    0022EE44 00402BE1 (0A077DA8, 00000000, 00000001,
    00000000)
    0022EE74 00419367 (0A077DA8, 00000001, 00002A8F,
    0003EC3E)
    0022EED4 00418993 (0A077DA8, 00000009, 00000000,
    00000000)
    0022EF14 0041774A (0A077DA8, 00000008, 0022EF74,
    00419661)
    0022EF64 00418CDE (0A077DA8, 00000007, 3EE20769,
    00000000)
    0022EFD4 00418283 (0A01ADD0, 00000007, 00002A8F,
    00094A7D)
    0022F014 0041774A (0A01ADD0, 00000006, 0022F074,
    0041961C)
    0022F064 00418CDE (0A01ADD0, 00000005, 00000001,
    00000000)
    0022F0D4 00418283 (0A01A830, 00000005, 00000000,
    00000000)
    0022F114 0041774A (0A01A830, 00000004, 0022F174,
    00419661)
    0022F164 00418CDE (0A01A830, 00000003, 00000001,
    00000000)
    End of stack trace (more stack frames may be present)

    ==============================================

    g++ -W -Wall -Wpointer-arith -mcpu=pentiumpro -mthreads -
    W -Wall -Wpointer-a
    rith -O3 -g -Wno-uninitialized -I/d/ACE_wrappers/ -
    DACE_HAS_EXCEPTIONS -
    DACE_RMCAST_BUILD_DLL -c -
    o .shobj/RMCast_UDP_Reliable_Sender.o
    RMCast_UDP_Reliable_Sender.cpp
    0 [main] make 1004 open_stackdumpfile: Dumping stack
    trace to
    make.exe.stackdump
    make[2]: *** [RMCast.subdir] Segmentation fault (core
    dumped)
    .................................................
    Exception: STATUS_ACCESS_VIOLATION at eip=71098C0D
    eax=00000000 ebx=00000000 ecx=000000D0 edx=00000000
    esi=0A0F1E18
    edi=0A0F1D48
    ebp=0022E934 esp=0022E90C
    program=C:\util\msys\bin\make.exe
    cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023
    Stack trace:
    Frame Function Args
    0022E934 71098C0D (710A5020, 0A0F1D50, 0022EE24,
    0041C94B)
    0022E954 7103110D (0A0F1D50, 0041C4E2, 0A010AA0,
    00000198)
    0022EE24 0041C964 (0A01ADD0, 00000000, 0A0F1B9A,
    0A0739D4)
    0022EE84 0040C23C (0A0739B8, 71097E93, 0022EEC4,
    00404647)
    0022EEC4 0040C51E (0A0739B8, 0A01ADD0, 0022EF14,
    0040C69F)
    0022EF14 0040C9A4 (0A01ADD0, 00000000, 00000001,
    00419661)
    0022EF44 00402BE1 (0A01ADD0, 00000000, 3EE20769,
    00000000)
    0022EF74 00419367 (0A01ADD0, 00000001, 0A01B010,
    0040A895)
    0022EFD4 00418993 (0A01ADD0, 00000007, 00002A8F,
    00094A7D)
    0022F014 0041774A (0A01ADD0, 00000006, 0022F074,
    0041961C)
    0022F064 00418CDE (0A01ADD0, 00000005, 00000001,
    00000000)
    0022F0D4 00418283 (0A01A830, 00000005, 00000000,
    00000000)
    0022F114 0041774A (0A01A830, 00000004, 0022F174,
    00419661)
    0022F164 00418CDE (0A01A830, 00000003, 00000001,
    00000000)
    0022F1D4 00418283 (0A01AB58, 00000003, 00000000,
    00000000)
    0022F214 0041774A (0A01AB58, 00000002, 0022F274,
    00419661)
    End of stack trace (more stack frames may be present)

    ==============================================

    g++ -W -Wall -Wpointer-arith -mcpu=pentiumpro -mthreads -
    W -Wall -Wpointer-a
    rith -O3 -g -Wno-uninitialized -I/d/ACE_wrappers/ -
    DACE_HAS_EXCEPTIONS
    -I/d/ACE_wrappers//TAO -DTAO_VALUETYPE_BUILD_DLL -
    c -o
    .shobj/ValueFactory.o ValueFactory.cpp
    0 [main] make 3304 open_stackdumpfile: Dumping stack
    trace to
    make.exe.stackdump
    make[2]: *** [Valuetype.subdir] Segmentation fault (core
    dumped)
    ...................................................
    Exception: STATUS_ACCESS_VIOLATION at eip=71098C0D
    eax=00000000 ebx=00000000 ecx=00000080 edx=00000000
    esi=0A0DEF50
    edi=0A0DEED0
    ebp=0022ED14 esp=0022ECEC
    program=C:\util\msys\bin\make.exe
    cs=001B ds=0023 es=0023 fs=0038 gs=0000 ss=0023
    Stack trace:
    Frame Function Args
    0022ED14 71098C0D (710A5020, 0A0DEF48, 00000001,
    00000000)
    0022ED34 7103110D (0A0DEF48, 00080000, 00000000,
    0022ED74)
    0022ED74 0040BEEE (0A0DE148, 0022EDC0, 0022EDC4,
    0040BCEC)
    0022EDC4 0040BD20 (00000001, 00000000, 0022EE14,
    0040C69F)
    0022EE14 0040C9CF (0A068988, 00000000, 00000000,
    0041961C)
    0022EE44 00402BE1 (0A068988, 00000000, 00000001,
    00000000)
    0022EE74 00419367 (0A068988, 00000001, 00002A8F,
    0003EDCF)
    0022EED4 00418993 (0A068988, 00000009, 00000000,
    00000000)
    0022EF14 0041774A (0A068988, 00000008, 0022EF74,
    00419661)
    0022EF64 00418CDE (0A068988, 00000007, 00000001,
    00000000)
    0022EFD4 00418283 (0A01BE50, 00000007, 00000000,
    00000000)
    0022F014 0041774A (0A01BE50, 00000006, 0022F074,
    00419661)
    0022F064 00418CDE (0A01BE50, 00000005, 00000001,
    00000000)
    0022F0D4 00418283 (0A01BA38, 00000005, 00000000,
    00000000)
    0022F114 0041774A (0A01BA38, 00000004, 0022F174,
    00419661)
    0022F164 00418CDE (0A01BA38, 00000003, 00000001,
    00000000)
    End of stack trace (more stack frames may be present)

    </STACKDUMP>

     
  • Earnie Boyd
    Earnie Boyd
    2003-06-10

    Logged In: YES
    user_id=15438

    Ok, now this is a problem in make itself. It's dieing
    trying to free memory allocated to pass the child an
    environment. Does your process use multiple processors via
    the -j switch?

    If you want to help debug, this blew at job.c:825 of the
    make package. If you need the source, check out the MinGW
    CVS msys/packages/make/3.79.1 module.

    Earnie.

     
  • Logged In: YES
    user_id=585332

    We don't use multiple processors and so far the person that
    gave me the stackdumps also not.

     
  • Justin Forest
    Justin Forest
    2003-06-11

    Logged In: YES
    user_id=664367

    The latest snapshot is much better, still it's not perfect.
    It crashed on me twice already, though after all the build
    steps have already finished. Stackdump and msysinfo output
    are available at
    http://hex.forestnet.org/files/make-crash-20030611.txt

     
1 2 > >> (Page 1 of 2)