/g/ - syscall of the day: set_tid_address - Technology

Anonymous

syscall of the day: set_tid_ad(...) 05/10/26(Sun)15:56:37 No.108794806

File: huabqu3dfm6e1.jpg (146 KB, 1080x1197)

syscall of the day: set_tid_address Anonymous 05/10/26(Sun)15:56:37 No.108794806 Archived

previous: >>108786358
#define __NR_set_tid_address        218
https://man7.org/linux/man-pages/man2/set_tid_address.2.html

tl;dr:
manage a thread's tid

kind of a weird/niche one. it seems to be used in what is effectively a waitpid, but for threads?
why is returns your own tid, i can't even begin to imagine
it's a bit strange that this syscall supposedly always succeeds. i haven't looked at the source code, but i have a feeling if the address you pass in is outside of your address space...
just looked it up, and nope. lol
            /*
             * We don't check the error code - if userspace has
             * not set up a proper pointer then tough luck.
             */
relevant resources:
man man
man syscalls
https://man7.org/linux/man-pages/
https://linux.die.net/man/
https://elixir.bootlin.com/linux/
https://elixir.bootlin.com/musl/
https://elixir.bootlin.com/glibc/

Anonymous
05/10/26(Sun)17:16:29 No.108795244

Anonymous 05/10/26(Sun)17:16:29 No.108795244

File: 1767214649302229.png (164 KB, 407x485)

164 KB PNG

Usecase for altering a thread's ID?
Unironically.

Anonymous
05/10/26(Sun)17:38:29 No.108795364

Anonymous 05/10/26(Sun)17:38:29 No.108795364

very convoluted naming and manpage.
I wonder whether it's this convoluted just so can wait on a futex instead of always having to issue a syscall like with waitpid

Anonymous
05/10/26(Sun)17:39:55 No.108795373

Anonymous 05/10/26(Sun)17:39:55 No.108795373

>>108795244
you aren't altering the TID, the naming is just very confusing

Anonymous
05/10/26(Sun)19:07:48 No.108795873

Anonymous 05/10/26(Sun)19:07:48 No.108795873

>>108795364
glibc uses it for pthread joining.
https://elixir.bootlin.com/glibc/glibc-2.43/source/nptl/pthread_create.c#L293