In message <_A3373@delegate-en.ML_> on 07/07/06(18:34:40)
you "Xavier Cheney" <firstname.lastname@example.org> wrote:
| I had the windows bug. The window ask me to debug (but I haven't VC++ and
|it makes nothing when I click on debug) or to see/send error report. The
|error report indicates : szAppName : delegated.exe szAppVer :
|0.0.0.0 szModName : unknown szModVer : 0.0.0.0 offset : 00000000
| The technical report is stored on my computer and here is the content :
| I don't understand why there's MSXSL. I use it before launching delegate
|(to generate myfilter.php). File is stored in C:\delegate\admin\msxsl.
|Perhaps it's an old report, or perhaps because the file is stored in
| Here is the delegate log. You'll see the time between 11:07 and 11:09,
|time were I had the Windows bug (finishing by clicking
| I hope it'll help you ...
I feel I'm getting so closer to the bug :)
It seems that the DeleGate is jumping to somewhere unknown, possibly to
address zero. The strange report about MSXSL might be the result of it.
And I have a "good" news for you :)
I have found a path to reproduce the error constantly.
1) MOUNT the error message (for 403-forbidden)
2) cause repetitieve errors (getting forbidden responses)
3) then DeleGate causes "penalty delay"
4) then a socket for accepting clients is closed. possibly something is
5) MOUNTed error message from a remote server is got. something is broken
here, maybe it's the socket environment to be inherited on spawn.
6) client send the next request from the same connection on which it
was rejected. (the connection is kept alive in this case)
7) then CFI process is spawned. the connection between it and the caller
(parent) DeleGate process is broken, receiving empty environment,
including broken jump-address (maybe zero) to a function.
8) jumping to broken address causes SIGSEGV.
I'll inspect what happens in the phase 4-5) and fix it in the next release.
9 9 Yutaka Sato <email@example.com> http://delegate.org/y.sato/
( ~ ) National Institute of Advanced Industrial Science and Technology
_< >_ 1-1-4 Umezono, Tsukuba, Ibaraki, 305-8568 Japan
Do the more with the less -- B. Fuller