RE:[Bacula-Users] Baculs 3.31 crashes on restore

[Marc B. d. M. <bru...@is...>]

On 15 Jul 2003 at 11:08, Alessandro Bianchi wrote:

Hello Alessandro,

I have had the same problem and Kern send me a patch.
After applying the patch the restore worked.
You will find it attachend to this eMail

Here is the original mail from Kern:

Hello Marc,

Sorry for the problems. Sometimes it frustrates me because
Linux seems to be very forgiving on bad pointer references.
This never showed up here.  From the dump, I found the problem
(a simple missing pointer reference).

If you apply the following patch, it will fix the problem
(I hope):

  cd <bacula-source>/src/dird
  patch -p0 <tree.patch

Please let me know.

Best regards,

Kern




> Hi Kern
> 
> Still following your wonderful bacula, but got a problem with the new
> 1.31 release.
> 
> Whenever I try to start a restore , after selecting the job I get the
> following traceback: [New Thread 1075422528 (LWP 8823)] Error while
> reading shared library symbols: Can't attach LWP 8823: No such process
> 0xffffe002 in ?? () $1 = "skynet-dir", '\0' <repeats 19 times> $2 =
> 0x80a2f98 "bacula-dir" $3 = 0x80a2fc0 "/sbin/bacula-dir" $4 = "MySQL" $5
> = 0x8090343 "1.31 (10 Jul 2003)"
> #0  0xffffe002 in ?? ()
> #1  0x08057ad7 in wait_for_next_job(char*) (job_to_run=0xbfffea28 "<")
>      at scheduler.c:98
> #2  0x0804b047 in main (argc=0, argv=0x1) at dird.c:226
> #3  0x42015574 in __libc_start_main () from /lib/tls/libc.so.6
> [New Thread 1100639168 (LWP 8832)]
> [New Thread 1092250688 (LWP 8830)]
> [New Thread 1083862208 (LWP 8829)]
> 
> Thread 4 (Thread 1083862208 (LWP 8829)):
> #0  0xffffe002 in ?? ()
> #1  0x08063867 in connect_thread (arg=0x4) at ua_server.c:90
> #2  0x4002b2b6 in start_thread () from /lib/tls/libpthread.so.0
> 
> Thread 3 (Thread 1092250688 (LWP 8830)):
> #0  0xffffe002 in ?? ()
> #1  0x0807e62c in btimer_thread (arg=0x0) at watchdog.c:172
> #2  0x4002b2b6 in start_thread () from /lib/tls/libpthread.so.0
> 
> Thread 2 (Thread 1100639168 (LWP 8832)):
> #0  0xffffe002 in ?? ()
> #1  0x08078b5b in signal_handler (sig=11) at signal.c:138
> #2  <signal handler called>
> #3  new_ua_context(JCR*) (jcr=0x9090900d) at ua_server.c:199
> #4  0x08064bf2 in user_select_files_from_tree(TREE_CTX*)
> (tree=0x419a66cc)
>      at ua_tree.c:78
> #5  0x0805ff11 in restorecmd(UAContext*, char*) (ua=0x80a7ea8,
>      cmd=0x80a61a0 "5") at ua_restore.c:189
> #6  0x08057fc0 in do_a_command(UAContext*, char*) (ua=0x80a7ea8,
>      cmd=0x80a61a0 "5") at ua_cmds.c:144
> #7  0x08063a4f in handle_UA_client_request (arg=0x80a3068) at
> ua_server.c:155
> #8  0x0807f0a2 in workq_server (arg=0x80a2760) at workq.c:346
> #9  0x4002b2b6 in start_thread () from /lib/tls/libpthread.so.0
> /sbin/btraceback.gdb:7: Error in sourced command file:
> Couldn't get registers: No such process.
> #0  0xffffe002 in ?? ()
> 
> The problema appears exactely after bacula auto-selects the only client
> actually configured.
> 
> I compiled bacula from source (10 jul 2003) on RH 9.
> 
> Am I doing something wrong?
> 
> Thanks
> 
> Alessandro
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email sponsored by: Parasoft
> Error proof Web apps, automate testing & more.
> Download & eval WebKing and get a free book.
> www.parasoft.com/bulletproofapps
> _______________________________________________
> Bacula-users mailing list
> Bac...@li...
> https://lists.sourceforge.net/lists/listinfo/bacula-users
> 

-- 
Institut für Seeverkehrswirtschaft und Logistik	http://www.isl.org/
Dipl.-Ing. Marc Brückner                                 	
mailto:bru...@is...
Universitaetsallee GW1 Block A                      	phone:+49 421 
22096-67
D-28359 Bremen                                    	fax:+49 421 22096-55




-------------- Enclosure number 1 ----------------
--- ua_tree.c.old	Mon Jul 14 11:42:26 2003
+++ ua_tree.c	Mon Jul 14 11:42:01 2003
@@ -75,7 +75,7 @@
 {
    char cwd[2000];
    /* Get a new context so we don't destroy restore command args */
-   UAContext *ua = new_ua_context(ua->jcr);
+   UAContext *ua = new_ua_context(tree->ua->jcr);
    ua->UA_sock = tree->ua->UA_sock;   /* patch in UA socket */
 
    bsendmsg(tree->ua, _(