[Savannah-hackers] Re: lsh died on subversions.gnu.org

[Niels Möller]

Loic Dachary <address@hidden> writes:

>       lsh died on subversions.gnu.org. Here are the last syslog lines:

...

> Dec 27 06:22:09 subversions lshd[867]: io.c: Preparing fd 10 for reading and 
> writing 
> Dec 27 06:22:09 subversions lshd[867]: Root /cvsroot/sdx 
> Dec 27 06:22:09 subversions lshd[867]: read_line.c:100: do_read_line: Raising 
> exception Line too long. (type 4096), using handler installed by 
> connection.c:276: make_ssh_connection 
> Dec 27 06:22:09 subversions lshd[867]: connection.c:208: 
> do_exc_connection_handler: Raising exception Stop reading (type 65538), using 
> handler installed by handshake.c:355: do_handshake 
> Dec 27 06:22:09 subversions lshd[867]: Valid-responses ok error 
> Valid-requests Checked-in New-entry Checksum Copy-file Updated Created 
> Update-existing Merged Patched Rcs-diff Mode Mod-time Removed Remove-entry 
> Set-static-directory Clear-static-directory Set-sticky Clear-sticky Template 
> Set-checkin-prog Set-update-prog Notified Module-expansion Wrapper-rcsOption 
> M Mbinary E F MT 

I'm guessing this is a plain cvs-over-rsh client that tries to talk to
the ssh port?

> Dec 27 06:22:09 subversions lshd[867]: read_buffered: fd died, 15 buffered 
> bytes discarded 
> Dec 27 06:22:09 subversions lshd[867]: io.c: Marking fd 10 for closing. 
> Dec 27 06:22:09 subversions lshd[867]: Connection died. 
> Dec 27 06:22:09 subversions lshd[867]: io.c: Closing fd 10. 
> Dec 27 06:22:36 subversions lshd[867]: io.c: accept on fd 10 
> Dec 27 06:22:36 subversions lshd[867]: Accepting connection from 
> 199.232.39.35, port 58288 
> Dec 27 06:22:36 subversions lshd[867]: Initiating handshake with 
> 199.232.39.35 
> Dec 27 06:22:36 subversions lshd[867]: io.c: Preparing fd 10 for reading and 
> writing 
> Dec 27 06:22:36 subversions lshd[867]: Garbage collecting while idle... 

Looks like it crashes in the gc. Interesting, I haven't seen any
gc-related bugs in a while. A gdb session with the core file, when you
get one, should tell us what object it is that is corrupted, and how.

>       I guess this bug is some kind of xmas present to you ;-)

Well... I hope we can catch it soon. At least, I'm back on the net
now, although I'll be somewhat busy over the weekend.

Regards,
/Niels