monotone-devel
[Top][All Lists]
Advanced

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

Re: [Monotone-devel] branch_leaves_sync_bug


From: Stephen Leake
Subject: Re: [Monotone-devel] branch_leaves_sync_bug
Date: Wed, 26 May 2010 04:07:08 -0400
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/23.1 (windows-nt)

Thomas Keller <address@hidden> writes:

> On some of the build bots the aforementioned tests fails in line 47.
> While it runs through for me locally, I still have seen errornous output
> from it which does not influence the checked return code:
>
> Locally (exit code 0):

This is a Mac, right?

> mtn: setting default server to file:///../branch_leaves_sync_bug/beth.db
> mtn: setting default branch include pattern to '*'
> mtn: setting default branch exclude pattern to ''
> mtn: setting default include pattern for server
> 'file:///../branch_leaves_sync_bug/beth.db' to '*'
> mtn: setting default exclude pattern for server
> 'file:///../branch_leaves_sync_bug/beth.db' to ''
> mtn: connecting to file:///../branch_leaves_sync_bug/beth.db
> mtn: finding items to synchronize:
> mtn: ticks: c="certificates"/256, k="keys"/1, r="revisions"/64
> mtn: ckr
> mtn: ticks: >="bytes in"/1024, <="bytes out"/1024, c="certs in"/3,
> C="certs out"/3, r="revs in"/1, R="revs out"/1
> mtn: ><cCrRC<
> mtn: successful exchange with
> file:///Users/tommyd/Entwicklung/monotone-trunk/tester_dir/tests/branch_leaves_sync_bug/beth.db
> mtn: fatal: error: network/session_base.cc:82: I(!is_pipe_pair())

This indicates Netxx is confused; it thinks the session is using a pipe
instead of a socket. That's a serious bug.

Since it's not changing the return code, it must be coming from the
spawned process. 

Aside: it would be really nice if we could change the "mtn:" prefix in
the spawned process via command line option, so we could tell which
process the messages were coming from.

> openSUSE build bot (exit code 1):
>
> mtn: setting default server to
> file:///../branch_leaves_sync_bug/beth.db

How did test.root get set to a relative path? I used it because I
thought it was always absolute.

Hmm. It's set in testlib.lua, by 'test.root = chdir(".")'. 'chdir' is
implemented in tester.cc, using 'get_current_working_dir'. Which is in
unix/fs.cc, which calls getcwd. [1] says:

    The getcwd() function shall place an absolute pathname of the
    current working directory in the array pointed to by buf 

So openSUSE is not in compliance with that standard. Not much we can do
about that.

[1] http://www.opengroup.org/onlinepubs/009695399/functions/getcwd.html

> mtn: connecting to file:///../branch_leaves_sync_bug/beth.db
> mtn: No such file or directory
> /path/to/mtn: fatal signal: Segmentation fault

mtn handles this cleanly when I get similar errors on Windows,
Debian, and RedHat. Sounds like openSUSE is just broken.

-- 
-- Stephe



reply via email to

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