[RFA] lengthen timeouts in gdb.base/checkpoint.exp

[RFA] lengthen timeouts in gdb.base/checkpoint.exp

[Doug Evans]

Hi.

I'm seeing timeout failures in gdb.base/checkpoint.exp
on the 7.1 branch.

Running ../../../../src/gdb/testsuite/gdb.base/checkpoint.exp ...
FAIL: gdb.base/checkpoint.exp: break2 with many checkpoints (timeout)
FAIL: gdb.base/checkpoint.exp: info checkpoints with at least 600 checkpoints (timeout)
FAIL: gdb.base/checkpoint.exp: kill all one (timeout)

Ok to apply to trunk and 7.1 branch?

I went with adding a number to the existing timeout value because
I didn't want to make any assumptions on what the existing timeout
value was (I can imagine wanting to globally increase the timeout
to some value when testing a slow serial target).
I cut-n-pasted the means to save/restore timeout from watchpoint.exp.

2010-03-10  Doug Evans  <dje@google.com>

	* gdb.base/checkpoint.exp: Fix comment.
	Lengthen timeout while doing >600 checkpoints test.
	Rename duplicate "kill all one" test to keep test names unique.

Index: checkpoint.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.base/checkpoint.exp,v
retrieving revision 1.11
diff -u -p -r1.11 checkpoint.exp
--- checkpoint.exp	1 Jan 2010 07:32:00 -0000	1.11
+++ checkpoint.exp	10 Mar 2010 23:46:20 -0000
@@ -328,7 +328,7 @@ gdb_test "restart 9" "Not found.*" "no m
 gdb_test "restart 10" "Not found.*" "no more checkpoint 10"
 
 #
-# Now let's try setting a large number of checkpoints (>1000)
+# Now let's try setting a large number of checkpoints (>600)
 #
 
 gdb_exit
@@ -352,6 +352,10 @@ gdb_expect {
     timeout { fail "(timeout) set checkpoint breakpoint" }
 }
 
+set prev_timeout $timeout
+set timeout [expr $timeout + 120]
+verbose "Timeout now $timeout sec."
+
 gdb_breakpoint $break2_loc
 gdb_test "continue" "breakpoint 2.*" "break2 with many checkpoints"
 
@@ -375,9 +379,13 @@ gdb_test_multiple "info checkpoints" $ms
 # OK, kill 'em all...
 #
 
-gdb_test "kill" "" "kill all one" \
+gdb_test "kill" "" "kill all one with many checkpoints" \
     "Kill the program being debugged.*y or n. $" "y"
 
+# Restore old timeout
+set timeout $prev_timeout
+verbose "Timeout now $timeout sec."
+
 #
 # Finished: cleanup
 #

Re: [RFA] lengthen timeouts in gdb.base/checkpoint.exp

[Michael Snyder]

Doug Evans wrote:
> Hi.
> 
> I'm seeing timeout failures in gdb.base/checkpoint.exp
> on the 7.1 branch.
> 
> Running ../../../../src/gdb/testsuite/gdb.base/checkpoint.exp ...
> FAIL: gdb.base/checkpoint.exp: break2 with many checkpoints (timeout)
> FAIL: gdb.base/checkpoint.exp: info checkpoints with at least 600 checkpoints (timeout)
> FAIL: gdb.base/checkpoint.exp: kill all one (timeout)
> 
> Ok to apply to trunk and 7.1 branch?
> 
> I went with adding a number to the existing timeout value because
> I didn't want to make any assumptions on what the existing timeout
> value was (I can imagine wanting to globally increase the timeout
> to some value when testing a slow serial target).
> I cut-n-pasted the means to save/restore timeout from watchpoint.exp.
> 
> 2010-03-10  Doug Evans  <dje@google.com>
> 
> 	* gdb.base/checkpoint.exp: Fix comment.
> 	Lengthen timeout while doing >600 checkpoints test.
> 	Rename duplicate "kill all one" test to keep test names unique.

I'm ok with the timeout part.  Are you really changing the number
of checkpoints, or was the comment already obsolete?


> Index: checkpoint.exp
> ===================================================================
> RCS file: /cvs/src/src/gdb/testsuite/gdb.base/checkpoint.exp,v
> retrieving revision 1.11
> diff -u -p -r1.11 checkpoint.exp
> --- checkpoint.exp	1 Jan 2010 07:32:00 -0000	1.11
> +++ checkpoint.exp	10 Mar 2010 23:46:20 -0000
> @@ -328,7 +328,7 @@ gdb_test "restart 9" "Not found.*" "no m
>  gdb_test "restart 10" "Not found.*" "no more checkpoint 10"
>  
>  #
> -# Now let's try setting a large number of checkpoints (>1000)
> +# Now let's try setting a large number of checkpoints (>600)
>  #
>  
>  gdb_exit
> @@ -352,6 +352,10 @@ gdb_expect {
>      timeout { fail "(timeout) set checkpoint breakpoint" }
>  }
>  
> +set prev_timeout $timeout
> +set timeout [expr $timeout + 120]
> +verbose "Timeout now $timeout sec."
> +
>  gdb_breakpoint $break2_loc
>  gdb_test "continue" "breakpoint 2.*" "break2 with many checkpoints"
>  
> @@ -375,9 +379,13 @@ gdb_test_multiple "info checkpoints" $ms
>  # OK, kill 'em all...
>  #
>  
> -gdb_test "kill" "" "kill all one" \
> +gdb_test "kill" "" "kill all one with many checkpoints" \
>      "Kill the program being debugged.*y or n. $" "y"
>  
> +# Restore old timeout
> +set timeout $prev_timeout
> +verbose "Timeout now $timeout sec."
> +
>  #
>  # Finished: cleanup
>  #

Re: [RFA] lengthen timeouts in gdb.base/checkpoint.exp

[Doug Evans]

On Wed, Mar 10, 2010 at 4:11 PM, Michael Snyder <msnyder@vmware.com> wrote:
> I'm ok with the timeout part.  Are you really changing the number
> of checkpoints, or was the comment already obsolete?

Yeah, the comment was obsolete.  I didn't change anything other than
lengthen the timeout for that part of the test.

Re: [RFA] lengthen timeouts in gdb.base/checkpoint.exp

[Daniel Jacobowitz]

On Wed, Mar 10, 2010 at 03:53:24PM -0800, Doug Evans wrote:
> Hi.
> 
> I'm seeing timeout failures in gdb.base/checkpoint.exp
> on the 7.1 branch.
> 
> Running ../../../../src/gdb/testsuite/gdb.base/checkpoint.exp ...
> FAIL: gdb.base/checkpoint.exp: break2 with many checkpoints (timeout)
> FAIL: gdb.base/checkpoint.exp: info checkpoints with at least 600 checkpoints (timeout)
> FAIL: gdb.base/checkpoint.exp: kill all one (timeout)
> 
> Ok to apply to trunk and 7.1 branch?

OK.

-- 
Daniel Jacobowitz
CodeSourcery