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 5/5] range stepping: tests


On 05/22/2013 03:02 PM, Yao Qi wrote:> On 05/21/2013 02:28 AM, Pedro Alves wrote:
>> I see.  Thanks, that's much more detailed info than just saying "it's unsafe".
>>
>> I'm guessing the huge number of RSP packets comes from that big loop in the
>> test:
>>
>>    /* Generate a range that includes a loop, which is time consuming.
>>       Variable C is used to terminate the loop earlier when GDB
>>       wants.  */
>>    for (c = 1, a = 0; a < 65535 && c; a++) {for (b = 0; b < 65535 && c; b++) { d1 = d2 * a / b; d2 = d1 *
>>
>> We could skip most of the range stepping tests if e.g., the
>> test that steps the short line FAILs:
>>
>>    /* A line of source will be generated to a number of
>>      instructions by compiler.  */
>>    a = b + c + d * e - a; /* location 1 */
>>
>> WDYT?
>
> Pedro,
> That is a good idea.  It works for my internal stub!  I'll post a delta patch on top of yours.

...

> The patch below is to skip the rest of range stepping tests if the
> first one fails, to avoid the issue of huge number of rsp packets in
> gdb.log.  It is on top of your series.  It works well with my stub.

Thanks!

> 2013-05-22  Yao Qi  <yao@codesourcery.com>
>
> 	* gdb.base/range-stepping.exp: Skip the rest of tests if test
> 	fails.
> ---
>  gdb/testsuite/gdb.base/range-stepping.exp |    8 ++++++++
>  1 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/gdb/testsuite/gdb.base/range-stepping.exp b/gdb/testsuite/gdb.base/range-stepping.exp
> index def25ce..aa5d34f 100644
> --- a/gdb/testsuite/gdb.base/range-stepping.exp
> +++ b/gdb/testsuite/gdb.base/range-stepping.exp
> @@ -85,6 +85,14 @@ with_test_prefix "multi insns" {
>  	    set pc_after_stepping $expect_out(1,string)
>  	    pass $msg
>  	}
> +	-re ".*" {

This should not leave the prompt in the buffer, as it may
confuse the next test.

But I'm not understanding how this is catching the issue.  Why
would "print/x \$pc" fail if the stub degenerates to
implementing vCont;r as a single instruction step?

> +	    fail $msg
> +	    # It is the first test on range-stepping, and the simplest
> +	    # one.  If it fails, probably the rest of the tests fail
> +	    # and huge number of rsp packets will blow up the gdb.log
> +	    # file.  Skip the rest of the tests.

Suggest minor editing:

   # This is the first range-stepping test, and the simplest
   # one.  If it fails, probably the rest of the tests would
   # fail too, and the huge number of rsp packets in the test with
   # the time-consuming loop would blow up the gdb.log file.
   # Skip the rest of the tests.

> +	    return
> +	}
>      }
>
>      # There should be at least two instructions between


-- 
Pedro Alves


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