chicken-janitors
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Chicken-janitors] [Chicken-bugs] #334: csi compiled without debugging s


From: Chicken Scheme
Subject: [Chicken-janitors] [Chicken-bugs] #334: csi compiled without debugging segfaults on NetBSD/amd64 3.0
Date: Fri, 05 Oct 2007 19:33:42 -0000

#334: csi compiled without debugging segfaults on NetBSD/amd64 3.0
---------------------+------------------------------------------------------
Reporter:  sjamaan   |       Owner:  felix         
    Type:  defect    |      Status:  new           
Priority:  critical  |   Component:  core libraries
 Version:  2.7       |    Keywords:                
---------------------+------------------------------------------------------
 I compiled the latest chicken (2.720, from today's trunk: r6292) with
 2.711 and 2.720, and csi segfaults upon startup.

 I tried to trace it, but it didn't really help a lot to do so because I
 had no debugging symbols.  Then I recompiled _with_ debugging symbols
 (gmake DEBUGBUILD=1 ...), and the problem went away!

 The trace showed some pretty weird stuff, too:
 {{{
   <set break on "C_intern_in" function, run csi>
   #0  C_intern_in (ptr=0x53f878, len=18, str=0xfa00 <Error reading address
 0xfa00: Invalid argument>, stable=0x200a4405a) at runtime.c:1869
   #1  0x0000000200a11386 in C_intern2 (ptr=0x53f878, str=0x12 <Error
 reading address 0x12: Invalid argument>) at runtime.c:1931
   <set break on "lookup" function, continue>
   (gdb) bt
   #0  lookup (key=2484, len=0, str=0x542040 "ÜC¤", stable=0x542040) at
 runtime.c:1958
   #1  0x0000000200a111f8 in C_intern_in (ptr=0x53f878, len=18,
 str=0x200a4405a "\003sysinterrupt-hook", stable=0x200a4405a) at
 runtime.c:1877
 }}}

 This points to some kind of stack shift, perhaps?  I got this weird stack
 corruption with the debug build, but NOT with a debug build which had -O0
 added to the C compiler flags.

 I compiled on NetBSD/macppc too, (4.0 RC1, a different version from the
 amd64 computer), but it didn't crash and the stack was not FUBAR in this
 way there.

 NetBSD/amd64 3.0 has GCC 3.3.3 and GDB 5.3

 NetBSD/macppc 4.0RC1 has GCC 4.1.2 and GDB 6.5

 Mario verified that the problem was not there on Ubuntu Linux i386 and
 amd64 with the following software:

 Linux/amd64 with kernel version 2.6.15-23-amd64-server has gcc 4.0.3

 Linux/x86 with kernel version 2.6.20-15-386 has gcc 4.1.2

 You can check this by using your Frohike account, if you want. (that's the
 amd64 machine I used)

-- 
Ticket URL: <http://trac.callcc.org/ticket/334>
Chicken Scheme <http://www.call-with-current-continuation.org/>
The CHICKEN Scheme-to-C compiler

reply via email to

[Prev in Thread] Current Thread [Next in Thread]