From: Andrew Burgess <aburgess@redhat.com>
To: Pedro Alves <pedro@palves.net>, gdb-patches@sourceware.org
Subject: Re: [PATCH 2/4] gdb.base/callfuncs.exp: Adjust for Windows
Date: Thu, 16 Jul 2026 22:26:02 +0100 [thread overview]
Message-ID: <87wluud3th.fsf@redhat.com> (raw)
In-Reply-To: <c47e7788-4886-4419-b6f9-5590c226d641@palves.net>
Pedro Alves <pedro@palves.net> writes:
> A bit of self-review...
>
> I did the same mistake I did recently, again. Here:
>
>> -if { ![prepare_for_testing "failed to prepare" $testfile $srcfile "$compile_flags additional_flags=-DPROTOTYPES"] } {
>> +if { ![prepare_for_testing "failed to prepare" $testfile-1 $srcfile "$compile_flags additional_flags=-DPROTOTYPES"] } {
>> perform_all_tests 1
>> }
>>
>> with_test_prefix "noproto" {
>> - if { ![prepare_for_testing "failed to prepare" $testfile $srcfile \
>> + if { ![prepare_for_testing "failed to prepare" $testfile-0 $srcfile \
>> "$compile_flags additional_flags=-DNO_PROTOTYPES"] } {
>> perform_all_tests 0
>> }
>
> Since rerun_and_prepare now always restart GDB, these prepare_for_testing calls
> can be turned into build_executable calls.
>
> I've done that now. The testcase still passes cleanly.
>
> From 974b26155dd4a128a0382b3b0b08d60d78f0a0de Mon Sep 17 00:00:00 2001
> From: Pedro Alves <pedro@palves.net>
> Date: Tue, 14 Jul 2026 19:48:30 +0100
> Subject: [PATCH] gdb.base/callfuncs.exp: Adjust for Windows
>
> On Windows, gdb.base/callfuncs.exp currently ends up skipping the
> "noproto" tests, like:
>
> UNTESTED: gdb.base/callfuncs.exp: noproto: failed to prepare
>
> This is because the testcase tries to overwrite the executable while
> GDB has the executable still open, which doesn't work on Windows:
>
> Executing on host: x86_64-w64-mingw32-gcc ... -o .../gdb.base/callfuncs/callfuncs
> .../x86_64-w64-mingw32/bin/ld.exe: cannot open output file .../gdb.base/callfuncs/callfuncs.exe: Permission denied
>
> Fix this with two changes:
>
> - Compile the prototyped and non-prototyped executables to two
> separate executables. This is just what we normally do, so that
> it's convenient to test manually against either of the executables.
>
> - Restart GDB instead of just re-running to main.
>
> Strictly speaking, either of the changes alone would fix it, but I
> think both are useful to do.
Is this true? After the first perform_all_tests call GDB is still
running and has the executable file open. Without the separate
filenames the second build_executable call will try to change the
executable that GDB holds open.
All you've done is split prepare_for_testing into build_executable and
clean_restart, but the order hasn't changed (and you call clean_restart
more often now).
Not that I object to these changes, I just disagree with the last
sentence as I don't think the clean_restart change alone will fix the
problem.
Reviewed-By: Andrew Burgess <aburgess@redhat.com>
Thanks,
Andrew
>
> Change-Id: I496e0da65f9d484079c5d9a1222104bec39ee98b
> ---
> gdb/testsuite/gdb.base/callfuncs.exp | 26 ++++++++++++++------------
> 1 file changed, 14 insertions(+), 12 deletions(-)
>
> diff --git a/gdb/testsuite/gdb.base/callfuncs.exp b/gdb/testsuite/gdb.base/callfuncs.exp
> index f2870c577f2..0ac1c603351 100644
> --- a/gdb/testsuite/gdb.base/callfuncs.exp
> +++ b/gdb/testsuite/gdb.base/callfuncs.exp
> @@ -334,12 +334,18 @@ proc fetch_all_registers {test} {
> # Global used by RERUN_AND_PREPARE to make test names unique.
> set rerun_count 0
>
> -proc rerun_and_prepare {} {
> +proc rerun_and_prepare {prototypes} {
> global rerun_count
>
> + clean_restart $::testfile-$prototypes
> +
> incr rerun_count
> with_test_prefix "rerun number ${rerun_count}" {
>
> + gdb_test_no_output "set print sevenbit-strings"
> + gdb_test_no_output "set print address off"
> + gdb_test_no_output "set width 0"
> +
> if { ![runto_main] } {
> return
> }
> @@ -359,11 +365,7 @@ proc rerun_and_prepare {} {
> }
>
> proc perform_all_tests {prototypes} {
> - gdb_test_no_output "set print sevenbit-strings"
> - gdb_test_no_output "set print address off"
> - gdb_test_no_output "set width 0"
> -
> - rerun_and_prepare
> + rerun_and_prepare $prototypes
>
> # Save all register contents.
> set old_reg_content \
> @@ -382,7 +384,7 @@ proc perform_all_tests {prototypes} {
> fail "gdb function calls preserve register contents"
> }
>
> - rerun_and_prepare
> + rerun_and_prepare $prototypes
> # Save all register contents.
> set old_reg_content \
> [fetch_all_registers "retrieve original register contents 2"]
> @@ -408,7 +410,7 @@ proc perform_all_tests {prototypes} {
> }
> }
>
> - rerun_and_prepare
> + rerun_and_prepare $prototypes
> # Set breakpoint at a function we will call from gdb.
> gdb_breakpoint add
> # Save all register contents.
> @@ -434,7 +436,7 @@ proc perform_all_tests {prototypes} {
> }
> }
>
> - rerun_and_prepare
> + rerun_and_prepare $prototypes
> # Set breakpoint at a function we will call from gdb.
> gdb_breakpoint add
> # Save all register contents.
> @@ -459,7 +461,7 @@ proc perform_all_tests {prototypes} {
> }
> }
>
> - rerun_and_prepare
> + rerun_and_prepare $prototypes
> # Set breakpoint at a function we will call from gdb.
> gdb_breakpoint add
> set old_reg_content \
> @@ -547,12 +549,12 @@ proc perform_all_tests {prototypes} {
>
> # Perform all tests with and without function prototypes.
>
> -if { ![prepare_for_testing "failed to prepare" $testfile $srcfile "$compile_flags additional_flags=-DPROTOTYPES"] } {
> +if { ![build_executable "failed to build" $testfile-1 $srcfile "$compile_flags additional_flags=-DPROTOTYPES"] } {
> perform_all_tests 1
> }
>
> with_test_prefix "noproto" {
> - if { ![prepare_for_testing "failed to prepare" $testfile $srcfile \
> + if { ![build_executable "failed to build" $testfile-0 $srcfile \
> "$compile_flags additional_flags=-DNO_PROTOTYPES"] } {
> perform_all_tests 0
> }
>
> base-commit: 490469846dcef89fe53668bdbba73591c64bed61
> prerequisite-patch-id: f9480ec6b27aac188161579fdec6f08d5065ac6f
> --
> 2.54.0
next prev parent reply other threads:[~2026-07-16 21:26 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-14 22:06 [PATCH 0/4] gdb/testsuite: "long double" infcall tests Pedro Alves
2026-07-14 22:06 ` [PATCH 1/4] gdb.base/nodebug.exp: Add long double testing Pedro Alves
2026-07-16 21:14 ` Andrew Burgess
2026-07-17 13:59 ` Pedro Alves
2026-07-22 18:41 ` Andrew Burgess
2026-07-14 22:06 ` [PATCH 2/4] gdb.base/callfuncs.exp: Adjust for Windows Pedro Alves
2026-07-14 22:25 ` Pedro Alves
2026-07-16 21:26 ` Andrew Burgess [this message]
2026-07-17 14:27 ` Pedro Alves
2026-07-22 18:42 ` Andrew Burgess
2026-07-14 22:06 ` [PATCH 3/4] gdb.base/callfuncs.c: factor out float/double functions Pedro Alves
2026-07-16 21:33 ` Andrew Burgess
2026-07-14 22:06 ` [PATCH 4/4] gdb.base/callfuncs.exp: Exercise "long double" Pedro Alves
2026-07-16 21:36 ` Andrew Burgess
2026-07-22 13:01 ` [PATCH 0/4] gdb/testsuite: "long double" infcall tests Pedro Alves
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87wluud3th.fsf@redhat.com \
--to=aburgess@redhat.com \
--cc=gdb-patches@sourceware.org \
--cc=pedro@palves.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox