This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH v2 2/3] PR remote/19496, interrupted syscall in forking-threads-plus-bkpt


Hi Don,

On 02/11/2016 12:28 AM, Don Breazeal wrote:

> So, the bottom line is that I haven't changed the fix for the interrupted
> system calls, because I can't find anything that works as well as the
> original fix.  Perhaps this test puts enough stress on the kernel that the
> kernel bugs mentioned above are exposed.

Many thanks for the thorough investigation.  Let's take your
original approach then.

> 
> One change I did make from the previous version was to increase the
> timeout to 90 seconds, which was necessary to get more reliable results
> on the Nios II target.
> diff --git a/gdb/testsuite/gdb.threads/forking-threads-plus-breakpoint.exp b/gdb/testsuite/gdb.threads/forking-threads-plus-breakpoint.exp
> index ff3ca9a..6889c2b 100644
> --- a/gdb/testsuite/gdb.threads/forking-threads-plus-breakpoint.exp
> +++ b/gdb/testsuite/gdb.threads/forking-threads-plus-breakpoint.exp
> @@ -73,6 +73,9 @@ proc do_test { cond_bp_target detach_on_fork displaced } {
>       global linenum
>       global is_remote_target
>   
> +    global timeout
> +    set timeout 90
> +

Use with_timeout_factor instead so that the timeout is properly restored,
and put it around the problematic test, only, instead of basically
around the whole test case.  I think that'll be the "inferior 1 exited"
test?

>       set saved_gdbflags $GDBFLAGS
>       set GDBFLAGS [concat $GDBFLAGS " -ex \"set non-stop on\""]
>       clean_restart $binfile

Thanks,
Pedro Alves


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]