Re: [PATCH v2 2/2] gdb/testsuite: Add test for AArch64 Scalable Vector Extension

[Thiago Jung Bauermann via Gdb-patches <gdb-patches@sourceware.org>]

Hello Luis,

Luis Machado <luis.machado@arm.com> writes:

> Hi,
>
> Sorry I didn't reply to the previous v1 discussion. See notes below.

No problem. Thank you for quickly reviewing the patches!

> On 8/5/22 01:46, Thiago Jung Bauermann via Gdb-patches wrote:
>> --- /dev/null
>> +++ b/gdb/testsuite/gdb.arch/aarch64-sve.exp
>> @@ -0,0 +1,101 @@
>> +# Copyright 2022 Free Software Foundation, Inc.
>> +
>> +# This program is free software; you can redistribute it and/or modify
>> +# it under the terms of the GNU General Public License as published by
>> +# the Free Software Foundation; either version 3 of the License, or
>> +# (at your option) any later version.
>> +#
>> +# This program is distributed in the hope that it will be useful,
>> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> +# GNU General Public License for more details.
>> +#
>> +# You should have received a copy of the GNU General Public License
>> +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
>> +
>> +# Test a binary that uses SVE and exercise changing the SVE vector length.
>> +
>> +if {[skip_aarch64_sve_tests]} {
>> +    verbose "Skipping ${gdb_test_file_name}."
>> +    return
>> +}
>> +
>> +standard_testfile
>> +if { [prepare_for_testing "failed to prepare" ${testfile} ${srcfile}] } {
>> +    return
>> +}
>> +
>> +set linespec ${srcfile}:[gdb_get_line_number "break here"]
>> +
>> +if ![runto ${linespec}] {
>> +    return
>> +}
>> +
>> +# Count number of lines in "info registers" output.
>> +proc count_info_registers {} {
>> +    global gdb_prompt
>> +    set ret 0
>> +
>> +    gdb_test_multiple "info all-registers" "" {
>
> The use of "info all-registers" seems to introduce a lot of verbosity to the test and make it much slower (as you pointed out).
>
> Checking the code, it seems we always have either FPU or SVE. So at first glance it may not be worth it to list all of the possible
> fpu/vector registers here.
>
> With that said, there are upcoming features that may add registers to the vector group (SME), so might as well check all the registers.
>
> That's a very verbose way to say the above looks good. :-)

Thank yo for the analysis. I think that the test completeness is worth
the extra time it takes.

>> diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
>> index a8f25b5f0dd5..b6bfa8eaafa5 100644
>> --- a/gdb/testsuite/lib/gdb.exp
>> +++ b/gdb/testsuite/lib/gdb.exp
>> @@ -7885,6 +7885,12 @@ proc multi_line_input { args } {
>>       return [join $args "\n"]
>>   }
>>   +# Return how many newlines there are in the given string.
>> +
>> +proc count_newlines { string } {
>> +    return [regexp -all "\n" $string]
>> +}
>> +
>>   # Return the version of the DejaGnu framework.
>>   #
>>   # The return value is a list containing the major, minor and patch version
>> diff --git a/gdb/testsuite/lib/mi-support.exp b/gdb/testsuite/lib/mi-support.exp
>> index ca56e12b06bf..e821c0f6914f 100644
>> --- a/gdb/testsuite/lib/mi-support.exp
>> +++ b/gdb/testsuite/lib/mi-support.exp
>> @@ -1728,10 +1728,6 @@ set mi_autotest_data ""
>>   # The name of the source file for autotesting.
>>   set mi_autotest_source ""
>>   -proc count_newlines { string } {
>> -    return [regexp -all "\n" $string]
>> -}
>> -
>>   # Prepares for running inline tests in FILENAME.
>>   # See comments for mi_run_inline_test for detailed
>>   # explanation of the idea and syntax.
>
> This looks good to me overall. I'd be more comfortable with someone else approving the generic bits (though they look
> quite obvious to me).

Ok, hopefully others will agree. :-)

> Again, thanks for the patch.

Thank you again for reviewing both versions of these patches!

-- 
Thiago