Re: fsfs: Segfault when rep line lists the all-zeroes checksum

2016-08-29 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Mon, Aug 29, 2016 at 22:10:07 +0200: > On 29.08.2016 18:57, Daniel Shahaf wrote: > >Line 801 sets CHECKSUM to NULL (as promised by svn_checksum_parse_hex()'s > >docstring), line 803 dereferences it unconditionally. > I vaguely remember that we use(d) all-0 checksums > as a

Re: fsfs: Segfault when rep line lists the all-zeroes checksum

2016-08-29 Thread Stefan Fuhrmann
On 29.08.2016 18:57, Daniel Shahaf wrote: When the "props:" line in a noderev lists an all-zeroes checksum, FSFS segfaults: [[[ % cat repro.sh #!/bin/sh rm -rf r wc svnadmin create r svn co -q file://`pwd`/r wc svn ps -q k v wc svn ci -qmm wc echo "First dump:" svnadmin dump -q r >/dev/null perl

Re: svn commit: r1758069 - /subversion/trunk/subversion/libsvn_subr/dirent_uri.c

2016-08-29 Thread Ivan Zhakov
On 29 August 2016 at 21:05, Stefan Sperling wrote: > On Mon, Aug 29, 2016 at 02:44:28PM +0200, Stefan Sperling wrote: >> If you have time to fix it, please do! > > I got an opportunity to fix this myself: http://svn.apache.org/r1758269 Great, thanks! I started looking to this issue, but then was d

Re: svn commit: r1758069 - /subversion/trunk/subversion/libsvn_subr/dirent_uri.c

2016-08-29 Thread Stefan Sperling
On Mon, Aug 29, 2016 at 02:44:28PM +0200, Stefan Sperling wrote: > If you have time to fix it, please do! I got an opportunity to fix this myself: http://svn.apache.org/r1758269

Re: Svn capability for SparkleShare

2016-08-29 Thread Paul Hammant
My fork of sparkleshare now does svn-up and svn-commit well enough, in anyone is interested. It's only working on the svn:// protocol, and even then only for allow-anon-write repos. - Paul Sent from my iPhone > On Aug 25, 2016, at 6:18 PM, Paul Hammant wrote: > > SparkleShare: https://www

fsfs: Segfault when rep line lists the all-zeroes checksum

2016-08-29 Thread Daniel Shahaf
When the "props:" line in a noderev lists an all-zeroes checksum, FSFS segfaults: [[[ % cat repro.sh #!/bin/sh rm -rf r wc svnadmin create r svn co -q file://`pwd`/r wc svn ps -q k v wc svn ci -qmm wc echo "First dump:" svnadmin dump -q r >/dev/null perl -pi -e 's/[0-9a-f]{16,}/"0" x length $&/e

Re: Unbounded memory usage in mod_dav + mod_headers/mod_deflate/...

2016-08-29 Thread Evgeny Kotkov
Evgeny Kotkov writes: > It might be possible to rework mod_dav_svn, although it's going to take > some time. Currently, every top-level handler receives an `ap_filter_t *` > and passes it further, and all these places would have to be updated so > that the actual writes would target r->output_fi

Re: svn commit: r1758069 - /subversion/trunk/subversion/libsvn_subr/dirent_uri.c

2016-08-29 Thread Stefan Sperling
On Mon, Aug 29, 2016 at 12:36:05PM +0300, Ivan Zhakov wrote: > On 28 August 2016 at 00:02, wrote: > > Author: stsp > > Date: Sat Aug 27 21:02:27 2016 > > New Revision: 1758069 > > > > URL: http://svn.apache.org/viewvc?rev=1758069&view=rev > > Log: > > Fix issue #4652 which shows how to trigger an

Re: svn commit: r1758069 - /subversion/trunk/subversion/libsvn_subr/dirent_uri.c

2016-08-29 Thread Ivan Zhakov
On 28 August 2016 at 00:02, wrote: > Author: stsp > Date: Sat Aug 27 21:02:27 2016 > New Revision: 1758069 > > URL: http://svn.apache.org/viewvc?rev=1758069&view=rev > Log: > Fix issue #4652 which shows how to trigger an assertion failure in > svn_dirent_get_absolute() by passing invalid input on

Re: Remove svn_client_ctx_t from svn_client_conflict_t

2016-08-29 Thread Ivan Zhakov
On 26 August 2016 at 13:36, Stefan Sperling wrote: > On Fri, Aug 26, 2016 at 01:15:16PM +0300, Ivan Zhakov wrote: >> Hi Stefan, >> >> Currently svn_client_conflict_t stores a pointer to svn_client_ctx_t >> inside and uses it for every operation like >> svn_client_conflict_tree_get_details() or >>