You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Jani Averbach <ja...@cc.jyu.fi> on 2002/10/17 16:31:54 UTC
Segfault recipe
Hi!
I really appreciate if somebody would verify this:
I have made this 3 times now by hand, and the result is allways same.
But if you ran it from script, it works!?!
If somebody would like test with exactly same data, I will mail it
privately because it is my physics labs. =)
#
# Here we go
#
cd ~/tmp/opiskelu/fys104/fys101-2
svn import -m '' http://localhost:2401/repos/test-2/ . opiskelu/fys104/fys101-2
cd ~/tmp
svn import -m '' http://localhost:2401/repos/test-2 opiskelu/
# here I made few look ups to repo with browser
svn import -m '' http://localhost:2401/repos/test-2 opiskelu/ opiskelu
mv opiskelu opiskelu-orig
svn co http://localhost:2401/repos/test-2/opiskelu
###
Result of last command (and print out of file system structure):
A opiskelu/fys104
A opiskelu/fys104/fys101-2
A opiskelu/fys104/fys101-2/kulma35.gnumeric
A opiskelu/fys104/fys101-2/kulma23.gnumeric
A opiskelu/fys104/fys101-2/G.tex
A opiskelu/fys104/fys101-2/kulma30.gnumeric
A opiskelu/fys104/fys101-2/v0.tex
A opiskelu/fys104/fys101-2/abstract.txt
A opiskelu/fys104/fys101-2/kulma35.tex
A opiskelu/fys104/fys101-2/kulma23.tex
A opiskelu/fys104/fys101-2/Makefile
A opiskelu/fys104/fys101-2/valokenno.ps
A opiskelu/fys104/fys101-2/tykki.ps
A opiskelu/fys104/fys101-2/kulma30.tex
A opiskelu/fys104/fys101-2/heittoliike.tex
A opiskelu/fys104/fys101-2/valokenno.sk
A opiskelu/fys104/fys101-2/tykki.sk
A opiskelu/fys104/fys101-2/G.gnumeric
A opiskelu/fys104/fys101-2/v0.gnumeric
Checked out revision 3.
Segmentation fault
Should I show gdb to it? What would be good break point?
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> > I note you said earlier that it works when run from a script, can you
> > go into more details. Do you mean that the exact same commands work
> > when run from a script?
>
> Yes.
Sounds like it might be your environment.
> > Which commands do you put in the script? How do you run the script?
> >
> I have an another terminal command's where I have log of my first command
> typing typing (cat svn-bug.txt), and I past them via X's copy paste to the
> second terminal -> segfault
>
> however If I run
>
> $ sh svn-bug.txt
Is sh your normal shell? I don't know the best way to check your
environment, try this: run printenv in the shell where you get a
segfault. Then run sh -c printenv in the shell where you run the
script. Do they show the same output?
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 17 Oct 2002, Philip Martin wrote:
> Do the Subversion regression tests pass?
For this build I can't swear at 100%, but same source and earlier build
with same build apr* pass the regression test.
> If they do pass, and they should, can you try
> repeating this checkout with the subversion executable in the build
> area.
>
I will, at the moment I try to build 3388. I will check if this is still
repeatable.
>
> I note you said earlier that it works when run from a script, can you
> go into more details. Do you mean that the exact same commands work
> when run from a script?
Yes.
> Which commands do you put in the script? How do you run the script?
>
I have an another terminal command's where I have log of my first command
typing typing (cat svn-bug.txt), and I past them via X's copy paste to the
second terminal -> segfault
however If I run
$ sh svn-bug.txt
everything went fine. svn-bug.txt is at the end of this email.
There is two differensies:
1. I run it from script
2. when I made it by hand, I made one loop up to repository via/with my
browser, just poke and see what there is
Immediately after that I execute that svn co ... which segfaults.
I think that if this 3388 segfaults, I made to the script few sleep's
and one lynx -dump URL, and see what hapens.
BR, Jani
$ cat svn-bug.txt
cd ~/tmp/opiskelu/fys104/fys101-2
svn import -m '' http://localhost:2401/repos/test-2/ . opiskelu/fys104/fys101-2
cd ~/tmp
# svn import -m '' http://localhost:2401/repos/test-2/opiskelu/fys104/fys101-2 opiskelu/
svn import -m '' http://localhost:2401/repos/test-2 opiskelu/
svn import -m '' http://localhost:2401/repos/test-2 opiskelu/ opiskelu
mv opiskelu opiskelu-orig
svn co http://localhost:2401/repos/test-2/opiskelu
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Bug, differencies between ra_local and ra_dav (Re: Segfault recipe)
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> $ /opt/svn/3388/bin/svn import -m '' file:///srv/svn/dbs/test-2/ . opiskelu/fys104/fys101-2
>
> subversion/libsvn_fs/tree.c:181: (apr_err=160013, src_err=0)
> svn: Filesystem has no item
> svn: file not found: transaction `1', path `opiskelu/fys104/fys101-2'
>
>
> $ /opt/svn/3388/bin/svn import -m '' file:///srv/svn/dbs/test-2/ . opiskelu/
> [snip away ok output]
>
> $ /opt/svn/3388/bin/svn ls file:///srv/svn/dbs/test-2/
> opiskelu/
>
>
> So it seems very likely, that ra_dav can't handle correctly or it
> handles different way that first commit, and late this causes segfault.
I don't think that's the cause of the segfault. The import over
ra_dav creates the URL http://host/repos/test-2/fys101-2/ by dropping
the opiskelu/fys104/ components. I think that's a bug, it ought to
fail just like the ra_local case, but I don't think it causes the
segfault.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Bug, differencies between ra_local and ra_dav (Re: Segfault recipe)
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On Fri, 18 Oct 2002, Michael Wood wrote:
>
> Just a thought. Does it make a difference if you access it from
> file:///path/to/repos instead of http://blah/?
>
$ /opt/svn/3388/bin/svn import -m '' file:///srv/svn/dbs/test-2/ . opiskelu/fys104/fys101-2
subversion/libsvn_fs/tree.c:181: (apr_err=160013, src_err=0)
svn: Filesystem has no item
svn: file not found: transaction `1', path `opiskelu/fys104/fys101-2'
$ /opt/svn/3388/bin/svn import -m '' file:///srv/svn/dbs/test-2/ . opiskelu/
[snip away ok output]
$ /opt/svn/3388/bin/svn ls file:///srv/svn/dbs/test-2/
opiskelu/
So it seems very likely, that ra_dav can't handle correctly or it
handles different way that first commit, and late this causes segfault.
I emphasize that my svn works otherwise ok. This is only case where it's
segfaulting.
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Michael Wood <mw...@its.uct.ac.za>.
Just a thought. Does it make a difference if you access it from
file:///path/to/repos instead of http://blah/?
--
Michael Wood <mw...@its.uct.ac.za>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On Fri, 18 Oct 2002, Swen Thuemmler wrote:
>
> Or perhaps the shell sets some limits? What does ulimit -a show?
>
$ ulimit -a
core file size (blocks) 0
data seg size (kbytes) unlimited
file size (blocks) unlimited
max memory size (kbytes) unlimited
stack size (kbytes) unlimited
cpu time (seconds) unlimited
max user processes 2048
pipe size (512 bytes) 8
open files 1024
virtual memory (kbytes) 4194302
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Swen Thuemmler <sw...@mediaways.net>.
On Thu, Oct 17, 2002 at 08:34:41PM +0100, Philip Martin wrote:
> Still looks like it must be your environment, LD_LIBRARY_PATH or
> LD_PRELOAD being the obvious candidates. Are you using shared
> libraries? What does running ldd on the svn executable show, do the
> library paths look like the ones you expect?
Or perhaps the shell sets some limits? What does ulimit -a show?
Just wild guessing...
--Swen
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 17 Oct 2002, Philip Martin wrote:
>
> Still looks like it must be your environment, LD_LIBRARY_PATH or
> LD_PRELOAD being the obvious candidates.
>
It seems that the fault was in my system, however I don't know where.
I rebuild and recheck everything, and now subversion is running cleanly
both ra_local and ra_dav regression test, and also that mine offensive
test.
I also build subversion in another machine and it was also ok.
Sorry for noise, and thanks for Philip Martin for testing my repo.
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: regression testing over ra_dav
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 18 Oct 2002, Philip Martin wrote:
>
> No, see subversion/tests/clients/cmdline/README.
Oh, shame on me.
I will made few test more with another system (os version + hardware)
during weekend. And I will run regression test over ra_dav. I will be back
on sunday evening or monday mornig (UTC).
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: regression testing over ra_dav
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> If I change following file_schema_prefix to http://localhost:2401
> 99 file_schema_prefix = 'file://'
> in /subversion/tests/clients/cmdline/svntest/main.py, should I get
> regression test running over ra_dav?
No, see subversion/tests/clients/cmdline/README. To run the tests over
ra_dav you run
$ nice make check BASE_URL=http://host
or
$ nice make check BASE_URL=http://host:port
or to run one set of tests
$ ./update_tests.py --url http://host:port
Make sure the URL doesn't have a trailing slash.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
regression testing over ra_dav
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 18 Oct 2002, Philip Martin wrote:
>
> That's how libtool works.
>
Ok, thanks!
If I change following file_schema_prefix to http://localhost:2401
99 file_schema_prefix = 'file://'
in /subversion/tests/clients/cmdline/svntest/main.py, should I get
regression test running over ra_dav?
I guess, that I have to change also some other path, expecial svnadmin
create, yes?
Is there out of box way to make these test over ra_dav?
If not, I would look at them.
BR, Jani
--
Jani Averbach jaa@iki.fi +358 40 759 0984
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> This is intresting:
> ldd ~/data/sw/svn_stuff/builddir/3388/subversion/clients/cmdline/svn
> =>> not a dynamic executable
>
> Why not? svn has been build with following configure:
That's how libtool works. That file is a shell script, it links an
executable against the build libraries. You will find the executable
in clients/cmdline/.libs/lt-svn, once you have run the script.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 17 Oct 2002, Philip Martin wrote:
>
> Still looks like it must be your environment, LD_LIBRARY_PATH or
> LD_PRELOAD being the obvious candidates.
Both are unset.
> Are you using shared libraries?
Yes
> What does running ldd on the svn executable show, do the
> library paths look like the ones you expect?
>
Yes, they seems very reasonable, same for ldconfig -p
This is intresting:
ldd ~/data/sw/svn_stuff/builddir/3388/subversion/clients/cmdline/svn
=>> not a dynamic executable
Why not? svn has been build with following configure:
./configure --prefix="/opt/svn/$(basename $(pwd))"
--enable-maintainer-mode
--with-apr=/opt/svnd/test-20021017-090406
--with-apr-util=/opt/svnd/test-20021017-090406
--with-apxs=/opt/svnd/test-20021017-090406/bin/apxs
ldd /opt/svn/3388/bin/svn
=>> dynamic
They both are segfaulting.
>
> I don't really know what else to suggest.
>
Is there any change that this test tiggers a bug? Svn seems otherwise
working ok. So I don't think that this is totally and only
about environment thing.
For example, I could made clean checkout with both of these segfaultin
svn-clients from svn.tigris.org... And I could made checkout from another
repos here ie. repo-1. Only if I feed new born repo with those script,
result is segfault when I try checkout.
Regression test ran cleanly in console where I managed produce
segfaults with those scripts.
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> Ok, I tried it with bash:
>
> bash ../svn-bug.txt -> Happy end, no failure.
> bash ../svn-bug3.txt -> Happy end, no failure
>
> cut'n paste one by one ../svn-bug3.txt's commands -> segfault
> cut'n paste whole ../svn-bug3.txt script -> segfault
>
> login to linux's virtual console:
> bash ../svn-bug3.txt > svn-bug3-` event.sh`.log 2>&1
>
> ==>>> BINGO!! Segfault!
Still looks like it must be your environment, LD_LIBRARY_PATH or
LD_PRELOAD being the obvious candidates. Are you using shared
libraries? What does running ldd on the svn executable show, do the
library paths look like the ones you expect?
I don't really know what else to suggest.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 17 Oct 2002, Philip Martin wrote:
Tested with version: 3388
> Do the Subversion regression tests pass?
>
Yes,
[21:25:35] ~/data/sw/svn_stuff/builddir/3388
jaa @ eowyn 311 $ make check
Running all tests in hashdump-test...success
Running all tests in path-test...success
Running all tests in random-test...success
Running all tests in stringtest...success
Running all tests in time-test...success
Running all tests in translate-test...success
Running all tests in svn-test2.sh...success
Running all tests in svn-test.sh...success
Running all tests in target-test.py...success
Running all tests in run-fs-tests.py...success
Running all tests in run-repos-tests.py...success
Running all tests in getopt_tests.py...success
Running all tests in basic_tests.py...success
Running all tests in commit_tests.py...success
Running all tests in update_tests.py...success
Running all tests in switch_tests.py...success
Running all tests in prop_tests.py...success
Running all tests in schedule_tests.py...success
Running all tests in log_tests.py...success
Running all tests in copy_tests.py...success
Running all tests in diff_tests.py...success
Running all tests in module_tests.py...success
Running all tests in merge_tests.py...success
Running all tests in stat_tests.py...success
Running all tests in trans_tests.py...success
Running all tests in svnadmin_tests.py...success
>
> can you try repeating this checkout with the subversion executable in
> the build area.
>
Result -> segfault
> Is sh your normal shell?
Noup, sorry. My normal shell is bash.
>
> Do they show the same output?
>
$ sh -c printenv > sh.env
$ printenv > my.env
$ diff sh.env my.env
18a19
> PS1=\n[\t] \w\n\u @ \h \# $
29,30c30
< _=/usr/bin/printenv
< SHLVL=3
---
> SHLVL=2
32a33
> _=/usr/bin/printenv
Ok, I tried it with bash:
bash ../svn-bug.txt -> Happy end, no failure.
bash ../svn-bug3.txt -> Happy end, no failure
cut'n paste one by one ../svn-bug3.txt's commands -> segfault
cut'n paste whole ../svn-bug3.txt script -> segfault
login to linux's virtual console:
bash ../svn-bug3.txt > svn-bug3-` event.sh`.log 2>&1
==>>> BINGO!! Segfault!
I am out of words.
I attach a log file and svn-bug3.txt with that email.
BR, Jani
--
Jani Averbach
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 17 Oct 2002, Philip Martin wrote:
Ok, now if I run it from console, it works (made segfault).
Here is commands (No needs for sleep or lynx):
Next I will run regression test from console, and try see if it makes
difference.
set -x
echo start script
cd ~/tmp/opiskelu/fys104/fys101-2
~/data/sw/svn_stuff/builddir/3388/subversion/clients/cmdline/svn import -m '' http://localhost:2401/repos/test-2/ . opiskelu/fys104/fys101-2
cd ~/tmp
~/data/sw/svn_stuff/builddir/3388/subversion/clients/cmdline/svn import -m '' http://localhost:2401/repos/test-2 opiskelu/
~/data/sw/svn_stuff/builddir/3388/subversion/clients/cmdline/svn import -m '' http://localhost:2401/repos/test-2 opiskelu/ opiskelu
mv opiskelu opiskelu-orig
~/data/sw/svn_stuff/builddir/3388/subversion/clients/cmdline/svn co http://localhost:2401/repos/test-2/opiskelu
echo done
BR, Jani
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> Program received signal SIGSEGV, Segmentation fault.
> 0x4018abbe in svn_path_is_child (path1=0x80a0c60 "opiskelu", path2=0x2
> <Address 0x2 out of bounds>, pool=0x809f168)
> at subversion/libsvn_subr/path.c:622
> 622 if (path1[0] != '/' && path2[0] == '/')
> gdb) bt
> #0 0x4018abbe in svn_path_is_child (path1=0x80a0c60 "opiskelu", path2=0x2
> <Address 0x2 out of bounds>, pool=0x809f168)
> at subversion/libsvn_subr/path.c:622
Something looks really broken, the access baton hash appears to
contain junk. Do the Subversion regression tests pass? They use this
code extensively. If they do pass, and they should, can you try
repeating this checkout with the subversion executable in the build
area.
> #1 0x4003f3d8 in do_close (adm_access=0x80a0c40, preserve_lock=0) at
> subversion/libsvn_wc/lock.c:465
> #2 0x4003f5ad in svn_wc_adm_close (adm_access=0x80a0c40) at
> subversion/libsvn_wc/lock.c:510
> #3 0x4004d4cf in close_edit (edit_baton=0x8089ef0) at
> subversion/libsvn_wc/update_editor.c:1869
I note you said earlier that it works when run from a script, can you
go into more details. Do you mean that the exact same commands work
when run from a script? Which commands do you put in the script? How
do you run the script?
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Jani Averbach <ja...@cc.jyu.fi>.
On 17 Oct 2002, Philip Martin wrote:
> Yes, please. No need for a break point, just run it under gdb until
> it crashes, then do bt and show us the stack trace.
>
Sorry for line breaks.
I am on irc, and I have still process attached to gdb.
BR, Jani
(gdb) run co http://localhost:2401/repos/test-2/opiskelu
Starting program: /usr/bin/svn co
http://localhost:2401/repos/test-2/opiskelu
A opiskelu/fys104
A opiskelu/fys104/fys101-2
A opiskelu/fys104/fys101-2/kulma35.gnumeric
A opiskelu/fys104/fys101-2/kulma23.gnumeric
A opiskelu/fys104/fys101-2/G.tex
A opiskelu/fys104/fys101-2/kulma30.gnumeric
A opiskelu/fys104/fys101-2/v0.tex
A opiskelu/fys104/fys101-2/abstract.txt
A opiskelu/fys104/fys101-2/kulma35.tex
A opiskelu/fys104/fys101-2/kulma23.tex
A opiskelu/fys104/fys101-2/Makefile
A opiskelu/fys104/fys101-2/valokenno.ps
A opiskelu/fys104/fys101-2/tykki.ps
A opiskelu/fys104/fys101-2/kulma30.tex
A opiskelu/fys104/fys101-2/heittoliike.tex
A opiskelu/fys104/fys101-2/valokenno.sk
A opiskelu/fys104/fys101-2/tykki.sk
A opiskelu/fys104/fys101-2/G.gnumeric
A opiskelu/fys104/fys101-2/v0.gnumeric
Checked out revision 3.
Program received signal SIGSEGV, Segmentation fault.
0x4018abbe in svn_path_is_child (path1=0x80a0c60 "opiskelu", path2=0x2
<Address 0x2 out of bounds>, pool=0x809f168)
at subversion/libsvn_subr/path.c:622
622 if (path1[0] != '/' && path2[0] == '/')
gdb) bt
#0 0x4018abbe in svn_path_is_child (path1=0x80a0c60 "opiskelu", path2=0x2
<Address 0x2 out of bounds>, pool=0x809f168)
at subversion/libsvn_subr/path.c:622
#1 0x4003f3d8 in do_close (adm_access=0x80a0c40, preserve_lock=0) at
subversion/libsvn_wc/lock.c:465
#2 0x4003f5ad in svn_wc_adm_close (adm_access=0x80a0c40) at
subversion/libsvn_wc/lock.c:510
#3 0x4004d4cf in close_edit (edit_baton=0x8089ef0) at
subversion/libsvn_wc/update_editor.c:1869
#4 0x4015be4a in wrap_close_edit (edit_baton=0x8088288) at
subversion/libsvn_delta/default_editor.c:507
#5 0x4016e85f in svn_ra_dav__do_checkout (session_baton=0x8088228,
revision=-1, recurse=1, editor=0x8089eb0, edit_baton=0x8089ef0)
at subversion/libsvn_ra_dav/fetch.c:1339
#6 0x40019b21 in svn_client_checkout (notify_func=0x804bd50 <notify>,
notify_baton=0x8060230, auth_baton=0x8087ea8,
URL=0x80601a0 "http://localhost:2401/repos/test-2/opiskelu",
path=0x8060220 "opiskelu", revision=0xbffff804, recurse=1,
xml_src=0x0, pool=0x8087dd8) at
subversion/libsvn_client/checkout.c:111
#7 0x804af7f in svn_cl__checkout (os=0x805fdd0, baton=0xbffff804,
pool=0x805fc70) at subversion/clients/cmdline/checkout-cmd.c:132
#8 0x804e883 in main (argc=3, argv=0xbffff8f4) at
subversion/clients/cmdline/main.c:833
--
Jani Averbach
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Segfault recipe
Posted by Philip Martin <ph...@codematters.co.uk>.
Jani Averbach <ja...@cc.jyu.fi> writes:
> Should I show gdb to it? What would be good break point?
Yes, please. No need for a break point, just run it under gdb until
it crashes, then do bt and show us the stack trace.
--
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org