2023-04-28 15:27:08

by Mark Brown

Subject: [PATCH] tools/nolibc: Implement WTERMSIG()

Add an implementation of WTERMSIG() to allow nolibc programs to obtain the
signal that caused the exit of a process, and use it in WIFEXITED() since
the condition being checked there is that no signal was reported.

Signed-off-by: Mark Brown <[email protected]>
---
tools/include/nolibc/types.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/include/nolibc/types.h b/tools/include/nolibc/types.h
index fbbc0e68c001..eb0b9f1ffe18 100644
--- a/tools/include/nolibc/types.h
+++ b/tools/include/nolibc/types.h
@@ -79,8 +79,9 @@
#define LINUX_REBOOT_CMD_SW_SUSPEND 0xd000fce2

/* Macros used on waitpid()'s return status */
+#define WTERMSIG(status) ((status) & 0x7f)
#define WEXITSTATUS(status) (((status) & 0xff00) >> 8)
-#define WIFEXITED(status) (((status) & 0x7f) == 0)
+#define WIFEXITED(status) (WTERMSIG(status) == 0)

/* waitpid() flags */
#define WNOHANG 1

---
base-commit: 457391b0380335d5e9a5babdec90ac53928b23b4
change-id: 20230427-nolibc-wtermsig-95c5d064f949

Best regards,
--
Mark Brown <[email protected]>

2023-05-02 06:36:52

by Willy Tarreau

Subject: Re: [PATCH] tools/nolibc: Implement WTERMSIG()

Hi Mark,

On Fri, Apr 28, 2023 at 04:24:36PM +0100, Mark Brown wrote:
> Add an implementation of WTERMSIG() to allow nolibc programs to obtain the
> signal that caused the exit of a process, and use it in WIFEXITED() since
> the condition being checked there is that no signal was reported.

Thanks, but Thomas already submitted a similar one that's already queued
for 6.5 in Paul's rcu tree:

https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git/commit/?h=dev.2023.04.26b&id=8c934d4822c77818d53f5f4c948c3ace6fda1505

At least it further emphasizes the fact that the patch is needed ;-)

Thanks!
Willy