* [RFC][gdb/testsuite] Register test for each arch separately in register_test_foreach_arch
@ 2021-09-13 15:40 Tom de Vries
2021-09-20 16:01 ` Simon Marchi
0 siblings, 1 reply; 3+ messages in thread
From: Tom de Vries @ 2021-09-13 15:40 UTC (permalink / raw)
To: gdb-patches
Hi,
In gdb/disasm-selftests.c we have:
...
selftests::register_test_foreach_arch ("print_one_insn",
selftests::print_one_insn_test);
...
and we get:
...
$ gdb -q -batch -ex "maint selftest print_one_insn" 2>&1 \
| grep ^Running
Running selftest print_one_insn.
$
...
Change the semantics register_test_foreach_arch such that a version of
print_one_insn is registered for each architecture, such that we have:
...
$ gdb -q -batch -ex "maint selftest print_one_insn" 2>&1 \
| grep ^Running
Running selftest print_one_insn::A6.
Running selftest print_one_insn::A7.
Running selftest print_one_insn::ARC600.
...
$
...
This makes it f.i. possible to do:
...
$ gdb -q -batch a.out -ex "maint selftest print_one_insn::armv8.1-m.main"
Running selftest print_one_insn::armv8.1-m.main.
Self test failed: self-test failed at src/gdb/disasm-selftests.c:165
Ran 1 unit tests, 1 failed
...
Currently this runs into a bfd problem, filed at PR28336 - "bfd printable arch
names not unique" ( https://sourceware.org/bugzilla/show_bug.cgi?id=28336 ).
Any comments?
Thanks,
- Tom
[gdb/testsuite] Register test for each arch separately in register_test_foreach_arch
---
gdb/selftest-arch.c | 105 ++++++++++++++++++++++------------------------------
1 file changed, 45 insertions(+), 60 deletions(-)
diff --git a/gdb/selftest-arch.c b/gdb/selftest-arch.c
index 052daed9196..935459def27 100644
--- a/gdb/selftest-arch.c
+++ b/gdb/selftest-arch.c
@@ -17,6 +17,7 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. */
#include "defs.h"
+#include <functional>
#if GDB_SELF_TEST
#include "gdbsupport/selftest.h"
@@ -25,73 +26,57 @@
namespace selftests {
-/* A kind of selftest that calls the test function once for each gdbarch known
- to GDB. */
-
-struct gdbarch_selftest : public selftest
+static bool skip_arch (const char *arch)
{
- gdbarch_selftest (self_test_foreach_arch_function *function_)
- : function (function_)
- {}
-
- void operator() () const override
- {
- std::vector<const char *> arches = gdbarch_printable_names ();
- bool pass = true;
-
- for (const char *arch : arches)
- {
- if (strcmp ("fr300", arch) == 0)
- {
- /* PR 20946 */
- continue;
- }
- else if (strcmp ("powerpc:EC603e", arch) == 0
- || strcmp ("powerpc:e500mc", arch) == 0
- || strcmp ("powerpc:e500mc64", arch) == 0
- || strcmp ("powerpc:titan", arch) == 0
- || strcmp ("powerpc:vle", arch) == 0
- || strcmp ("powerpc:e5500", arch) == 0
- || strcmp ("powerpc:e6500", arch) == 0)
- {
- /* PR 19797 */
- continue;
- }
-
- QUIT;
-
- try
- {
- struct gdbarch_info info;
-
- info.bfd_arch_info = bfd_scan_arch (arch);
-
- struct gdbarch *gdbarch = gdbarch_find_by_info (info);
- SELF_CHECK (gdbarch != NULL);
-
- function (gdbarch);
- }
- catch (const gdb_exception_error &ex)
- {
- pass = false;
- exception_fprintf (gdb_stderr, ex,
- _("Self test failed: arch %s: "), arch);
- }
-
- reset ();
- }
-
- SELF_CHECK (pass);
- }
+ if (strcmp ("fr300", arch) == 0)
+ {
+ /* PR 20946 */
+ return true;
+ }
+
+ if (strcmp ("powerpc:EC603e", arch) == 0
+ || strcmp ("powerpc:e500mc", arch) == 0
+ || strcmp ("powerpc:e500mc64", arch) == 0
+ || strcmp ("powerpc:titan", arch) == 0
+ || strcmp ("powerpc:vle", arch) == 0
+ || strcmp ("powerpc:e5500", arch) == 0
+ || strcmp ("powerpc:e6500", arch) == 0)
+ {
+ /* PR 19797 */
+ return true;
+ }
+
+ return false;
+}
- self_test_foreach_arch_function *function;
-};
+/* Register a kind of selftest that calls the test function once for each
+ gdbarch known to GDB. */
void
register_test_foreach_arch (const std::string &name,
self_test_foreach_arch_function *function)
{
- register_test (name, new gdbarch_selftest (function));
+ std::vector<const char *> arches = gdbarch_printable_names ();
+ for (const char *arch : arches)
+ {
+ if (skip_arch (arch))
+ continue;
+
+ auto test_fn
+ = ([=] ()
+ {
+ struct gdbarch_info info;
+ info.bfd_arch_info = bfd_scan_arch (arch);
+ struct gdbarch *gdbarch = gdbarch_find_by_info (info);
+ SELF_CHECK (gdbarch != NULL);
+ function (gdbarch);
+ reset ();
+ });
+
+ std::string test_name
+ = name + std::string ("::") + std::string (arch);
+ register_test (test_name, test_fn);
+ }
}
void
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [RFC][gdb/testsuite] Register test for each arch separately in register_test_foreach_arch
2021-09-13 15:40 [RFC][gdb/testsuite] Register test for each arch separately in register_test_foreach_arch Tom de Vries
@ 2021-09-20 16:01 ` Simon Marchi
2021-09-20 18:07 ` Tom de Vries
0 siblings, 1 reply; 3+ messages in thread
From: Simon Marchi @ 2021-09-20 16:01 UTC (permalink / raw)
To: Tom de Vries, gdb-patches
On 2021-09-13 11:40 a.m., Tom de Vries via Gdb-patches wrote:
> Hi,
>
> In gdb/disasm-selftests.c we have:
> ...
> selftests::register_test_foreach_arch ("print_one_insn",
> selftests::print_one_insn_test);
> ...
> and we get:
> ...
> $ gdb -q -batch -ex "maint selftest print_one_insn" 2>&1 \
> | grep ^Running
> Running selftest print_one_insn.
> $
> ...
>
> Change the semantics register_test_foreach_arch such that a version of
> print_one_insn is registered for each architecture, such that we have:
> ...
> $ gdb -q -batch -ex "maint selftest print_one_insn" 2>&1 \
> | grep ^Running
> Running selftest print_one_insn::A6.
> Running selftest print_one_insn::A7.
> Running selftest print_one_insn::ARC600.
> ...
> $
> ...
>
> This makes it f.i. possible to do:
> ...
> $ gdb -q -batch a.out -ex "maint selftest print_one_insn::armv8.1-m.main"
> Running selftest print_one_insn::armv8.1-m.main.
> Self test failed: self-test failed at src/gdb/disasm-selftests.c:165
> Ran 1 unit tests, 1 failed
> ...
>
> Currently this runs into a bfd problem, filed at PR28336 - "bfd printable arch
> names not unique" ( https://sourceware.org/bugzilla/show_bug.cgi?id=28336 ).
>
> Any comments?
I think the idea is good. It's still possible to run the test for all
arches, since the argument to "maint selftest" can match just a subset
of the test name. So we don't lose anything. The problem with PR28336
makes it so that we try to register tests with duplicate names, and that
asserts?
Simon
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [RFC][gdb/testsuite] Register test for each arch separately in register_test_foreach_arch
2021-09-20 16:01 ` Simon Marchi
@ 2021-09-20 18:07 ` Tom de Vries
0 siblings, 0 replies; 3+ messages in thread
From: Tom de Vries @ 2021-09-20 18:07 UTC (permalink / raw)
To: Simon Marchi, gdb-patches
On 9/20/21 6:01 PM, Simon Marchi wrote:
>
>
> On 2021-09-13 11:40 a.m., Tom de Vries via Gdb-patches wrote:
>> Hi,
>>
>> In gdb/disasm-selftests.c we have:
>> ...
>> selftests::register_test_foreach_arch ("print_one_insn",
>> selftests::print_one_insn_test);
>> ...
>> and we get:
>> ...
>> $ gdb -q -batch -ex "maint selftest print_one_insn" 2>&1 \
>> | grep ^Running
>> Running selftest print_one_insn.
>> $
>> ...
>>
>> Change the semantics register_test_foreach_arch such that a version of
>> print_one_insn is registered for each architecture, such that we have:
>> ...
>> $ gdb -q -batch -ex "maint selftest print_one_insn" 2>&1 \
>> | grep ^Running
>> Running selftest print_one_insn::A6.
>> Running selftest print_one_insn::A7.
>> Running selftest print_one_insn::ARC600.
>> ...
>> $
>> ...
>>
>> This makes it f.i. possible to do:
>> ...
>> $ gdb -q -batch a.out -ex "maint selftest print_one_insn::armv8.1-m.main"
>> Running selftest print_one_insn::armv8.1-m.main.
>> Self test failed: self-test failed at src/gdb/disasm-selftests.c:165
>> Ran 1 unit tests, 1 failed
>> ...
>>
>> Currently this runs into a bfd problem, filed at PR28336 - "bfd printable arch
>> names not unique" ( https://sourceware.org/bugzilla/show_bug.cgi?id=28336 ).
>>
>> Any comments?
>
> I think the idea is good.
Great :)
> It's still possible to run the test for all
> arches, since the argument to "maint selftest" can match just a subset
> of the test name. So we don't lose anything.
Ack.
> The problem with PR28336
> makes it so that we try to register tests with duplicate names, and that
> asserts?
Yes, but that PR's fixed now, so that assert no longer triggers.
[ FWIW, we could make things less dramatic by not asserting but throwing
an exception in register_test, and handling the exception in
register_test_foreach_arch. But there's no immediate need atm, so I'll
leave it as is in this patch. ]
Thanks,
- Tom
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2021-09-20 18:07 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-09-13 15:40 [RFC][gdb/testsuite] Register test for each arch separately in register_test_foreach_arch Tom de Vries
2021-09-20 16:01 ` Simon Marchi
2021-09-20 18:07 ` Tom de Vries
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).