Re: [PATCH] linux-user/strace: Improve output of various syscalls

[Aleksandar Markovic]

> @@ -26,7 +26,7 @@
>  { TARGET_NR_afs_syscall, "afs_syscall" , NULL, NULL, NULL },
>  #endif
>  #ifdef TARGET_NR_alarm
> -{ TARGET_NR_alarm, "alarm" , NULL, NULL, NULL },
> +{ TARGET_NR_alarm, "alarm" , "%s(%d)", NULL, NULL },
>  #endif

Man page says:

unsigned int alarm(unsigned int seconds)

The sole argument is unsigned int - therefore "%d" should be "%u",
shouldn't it?

--------------------------------------------------

This is not a part of your changes, but appeared in your patch diff:

>  #ifdef TARGET_NR_epoll_create
>  { TARGET_NR_epoll_create, "epoll_create" , NULL, NULL, NULL },

>From man pages:

int epoll_create(int size);

So, this also belongs to the category "has only int-type parameter,'
and "%s(%d)" should be used, no?

---------------------------------------------------

>  #ifdef TARGET_NR_setresgid
> -{ TARGET_NR_setresgid, "setresgid" , NULL, NULL, NULL },
> +{ TARGET_NR_setresgid, "setresgid" , "%s(%u,%u,%u)", NULL, NULL },
>  #endif
>  #ifdef TARGET_NR_setresgid32
>  { TARGET_NR_setresgid32, "setresgid32" , NULL, NULL, NULL },
>  #endif

Why are you here correcting setresgid(), but leaving setresgid32()
intact, even though they have the same argument type pattern?

--------------------------------------------------

I have these objections, however, in general, I salute the patch, and
your efforts to improve QEMU linux-user strace, it is a quite useful
debug tool, and thanks for doing this! :)

Yours,
Aleksandar