[Top][All Lists]

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

assertion failed in rpctrace

From: Gaël Le Mignot
Subject: assertion failed in rpctrace
Date: Mon, 18 Nov 2002 20:18:11 +0100
User-agent: Gnus/5.090008 (Oort Gnus v0.08) Emacs/21.2 (i386-pc-linux-gnu)


        While debugging a home-made program with rpctrace, I've found 
a bug in rpctrace: an assertion failed: 
rpctrace: ../../utils/rpctrace.c:699: trace_and_forward: Assertion `info' 

        I've looked inside rpctrace code,  and it's in the mangling of
the message header  before forwarding it, where the  port right passed
to rewrite_right is  a an invalid port (port 0 in  fact). I'm not sure
how this can happen exactly, I guess it's a bug in my own program, but
rpctrace shouldn't abort() like that when runned on a buggy program.

        Another interesting point is that when I run rpctrace on this
test program, it triggers the zalloc bug in GNU Mach (always with GNU
Mach 1.9, and sometimes with GNU Mach 1.3).

        The  test  program does  only  open()  and  then read()  on  a
home-made translator. You can use cat instead of the test porgram, the
result is the same (assertion  failed in rpctrace). The source code of
the translator, the test program and a script the triggers the bug can
be found at  http://kilobug.free.fr/hurd/hook-bug.tar.gz (you need the
list library from http://kilobug.free.fr/code/gnulibs-0.1.tar.gz too).

        If someone that undestands more than me the inside of rpctrace
and Mach IPC could look at  this, I'll be thankfull. I'm still avaible
if someone needs any precision.

PS:  I think  I've found  a gdb  bug too,  since I  cannot run  gdb on
rpctrace, it hangs when I try to run rpctrace; but that's another bug,
I'll collect some informations and send another mail.

Gael Le Mignot "Kilobug" - kilobug@freesurf.fr - http://kilobug.free.fr
GSM         : (in France)   ICQ UIN   : 7299959
Fingerprint : 1F2C 9804 7505 79DF 95E6 7323 B66B F67B 7103 C5DA

Member of HurdFr: http://hurdfr.org - The GNU Hurd: http://hurd.gnu.org

reply via email to

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