Re: [PATCH v3] [gdb] Create script to convert old tests into Dwarf::assemble calls.

[William Ferreira <wqferr@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2375 bytes --]

Copyright is already assigned, everything is ready. I can convert the tests
in a following commit, along with a few others.

I will have some free time starting on the 27th, so if you have any other
comments please submit a proper bug report before then. If you do so,
please send its URL to me either directly or in this thread.

Regarding your concerns as to whether that section will work properly,
could you please point me to a DWARF assembler test that uses this new
syntax? I might be able to fix this implementation before the rebase.

Thank you,
William Ferreira.

On Wed, Oct 15, 2025, 17:50 Tom Tromey <tom@tromey.com> wrote:

> >>>>> William Ferreira <wqferr@gmail.com> writes:
>
> > PR testsuite/32261 requests a script that could convert old .S-based
> > tests (that were made before dwarf.exp existed) into the new
> > Dwarf::assemble calls in Tcl. This commit is an initial implementation
> > of such a script. Python was chosen for convenience, and only relies on
> > a single external library.
>
> Thank you for doing this.
>
> > The following .exp files have been tried in this way and their outputs
> > highly resemble the original:
> > - gdb.dwarf2/dynarr-ptr
> > - gdb.dwarf2/void-type
> > - gdb.dwarf2/ada-thick-pointer
> > - gdb.dwarf2/atomic-type
> > - gdb.dwarf2/dw2-entry-points (*)
> > - gdb.dwarf2/main-subprogram
>
> Are you planning to try to convert these tests?  If so that would be
> fantastic.  If not, that's totally fine, maybe we should file a bug for
> this.  It'd be nice to get rid of the old .s tests.  Anyway let me know.
>
> > Currently the script has the following known limitations:
> [...]
>
> FWIW I think basically any limitations are fine.  The DWARF assembler
> itself had many at the start, people tend to add things as needed.
>
> > +        s = lbrace
> > +        s += self.name + " "
> > +        s += self._format_value(offset_die_lookup)
>
> I somewhat suspect this output will not work now, because there were
> some recent changes to make attributes Tcl code rather than
> specially-parsed data.
>
> However that's yet another thing we can fix in situ.
>
> Approved-By: Tom Tromey <tom@tromey.com>
>
> I forget what your copyright situation is.  If you are all set up, we
> can land this.  If you plan to submit more gdb contributions, let me
> know and we can set up write-after-approval access for you.
>
> Tom
>

[-- Attachment #2: Type: text/html, Size: 3294 bytes --]