Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Jan Kratochvil <jan.kratochvil@redhat.com>
To: Siddhesh Poyarekar <siddhesh@redhat.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [PATCH] Fix core dump test case failure with a ulimit check in gdb.base/auxv.exp
Date: Mon, 19 Mar 2012 09:43:00 -0000	[thread overview]
Message-ID: <20120319094218.GB14010@host2.jankratochvil.net> (raw)
In-Reply-To: <20120317073755.GB25580@spoyarek.redhat.com>

On Sat, 17 Mar 2012 08:37:58 +0100, Siddhesh Poyarekar wrote:
> @@ -42,8 +42,19 @@ set coredir "${objdir}/${subdir}/coredir.[getpid]"
>  file mkdir $coredir
>  set core_works [expr [isnative] && ! [is_remote target]]
>  
> -# Run GDB on the test program up to where it will dump core.
> +global gdb_prompt

Such "global" is not needed when it is at the file level.


> +gdb_exit
> +gdb_start
> +send_gdb "shell ulimit -c\n"
> +gdb_expect {
> +    -re "0\r\n$gdb_prompt $" {
> +        warning "ulimit -c is 0, disabling core tests"
> +        set core_works 0
> +    }
> +}

Other testcases try to set 'ulimit -c 0' first, this only testcase does not.
That should be fixed.

Also dealing with ulimit should be limited to the case where $core_works
- that means only for local native runs.  In such case one could also just run
'ulimit' by TCL, not by GDB 'shell'.

And as 'ulimit -c unlimited' in this case would be complicated - or at least
ot well compatible across platforms, there should be IMO setrlimit in the .c
file instead.  It should be protected by #ifdef and the compilation should be
tried with and without it, as I guess setrlimit is also not well compatible
across platforms.


> +# Run GDB on the test program up to where it will dump core.
>  gdb_exit
>  gdb_start
>  gdb_reinitialize_dir $srcdir/$subdir
> @@ -51,11 +62,9 @@ gdb_load ${binfile}
>  gdb_test_no_output "set print sevenbit-strings"
>  gdb_test_no_output "set width 0"
>  
> -if {$core_works} {
> -    if {[gdb_test "cd $coredir" ".*Working directory .*" \
> -	     "cd to temporary directory for core dumps"]} {
> -	set core_works 0
> -    }
> +if {[gdb_test "cd $coredir" ".*Working directory .*" \
> +         "cd to temporary directory for core dumps"]} {
> +    set core_works 0
>  }

I do not see a reason for this change.


>  if { ![runto_main] } then {
> @@ -154,7 +163,7 @@ if {$core_works} {
>  	pass $test
>      } else {
>  	set core_works 0
> -	warning "can't generate a core file - core tests suppressed - check ulimit -c"
> +	warning "can't generate a core file - core tests suppressed"
>  	fail $test

With the suggested 'ulimit -c unlimited' change to match other testcases
I think this change should be dropped.


>      }
>  } else {



Thanks,
Jan


  reply	other threads:[~2012-03-19  9:43 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-03-16 14:29 [PATCH] Fix gdb.base/auxv and gdb.base/cond-eval-mode test cases Siddhesh Poyarekar
2012-03-16 15:50 ` Pedro Alves
2012-03-17  7:27   ` Siddhesh Poyarekar
2012-03-19  9:35     ` [commit] " Jan Kratochvil
2012-03-17  7:37   ` [PATCH] Fix core dump test case failure with a ulimit check in gdb.base/auxv.exp Siddhesh Poyarekar
2012-03-19  9:43     ` Jan Kratochvil [this message]
2012-03-19 15:57       ` Siddhesh Poyarekar
2012-03-19 16:13         ` Jan Kratochvil
2012-03-19 16:39           ` Siddhesh Poyarekar
2012-03-19 18:40             ` [commit] " Jan Kratochvil

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120319094218.GB14010@host2.jankratochvil.net \
    --to=jan.kratochvil@redhat.com \
    --cc=gdb-patches@sourceware.org \
    --cc=siddhesh@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox