unofficial mirror of libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Szabolcs Nagy via Libc-alpha <libc-alpha@sourceware.org>
To: XingLi <lixing@loongson.cn>
Cc: adhemerval.zanella@linaro.org, libc-alpha@sourceware.org,
	i.swmail@xen0n.name, xry111@xry111.site, caiyinyu@loongson.cn,
	wanghongliang@loongson.cn, hejinyang@loongson.cn
Subject: Re: [PATCH] linux: Change syscall return value to long int
Date: Tue, 29 Nov 2022 08:55:24 +0000	[thread overview]
Message-ID: <Y4XI/A3whKF8BbA2@arm.com> (raw)
In-Reply-To: <20221129031659.2263453-1-lixing@loongson.cn>

The 11/29/2022 11:16, XingLi wrote:
> From: Xing Li <lixing@loongson.cn>
> 
> The kernel syscall return is long value.
> The generic syscall interface return value
> is int, which may lead to incorrect return value.

it's not clear what you mean here, the generic syscall
function returns long (according to unistd.h).

> 
> The following test is syscall with mmap executed on LoongArch,
> only 32bits and sign extension value returned leading to mmap failure,
> which should be with 47bits address returned.
> 
> Testcase:
> 
>  #include <sys/syscall.h>
>  #include <sys/mman.h>
>  #include <stdio.h>
> 
> void main()
> {
>         long int ret;
>         ret = syscall(SYS_mmap, NULL, 0x801000, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0);

Note: there are many reasons why direct calls to syscall
may not work.

syscall is a variadic argument function that takes long
arguments, but you pass ints that may *not* be sign/zero
extended on the caller site so e.g. the top 32bits of
size and offset can be arbitrary on a 64bit system.
(you have to cast args to long to make the example valid).

and some systems use SYS_mmap2.

>         printf("map address is %lx\n",ret);
> }
> 
> Result:
> [lixing@Sunhaiyong test]$ ./mmap
> map address is fffffffff008c000
> ---
>  sysdeps/unix/sysv/linux/syscall.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/sysdeps/unix/sysv/linux/syscall.c b/sysdeps/unix/sysv/linux/syscall.c
> index 7303ba7188..8cb0b66b1c 100644
> --- a/sysdeps/unix/sysv/linux/syscall.c
> +++ b/sysdeps/unix/sysv/linux/syscall.c
> @@ -33,7 +33,7 @@ syscall (long int number, ...)
>    long int a5 = va_arg (args, long int);
>    va_end (args);
>  
> -  int r = INTERNAL_SYSCALL_NCS_CALL (number, a0, a1, a2, a3, a4, a5);
> +  long int r = INTERNAL_SYSCALL_NCS_CALL (number, a0, a1, a2, a3, a4, a5);
>    if (__glibc_unlikely (INTERNAL_SYSCALL_ERROR_P (r)))

this change looks reasonable to me.

>      {
>        __set_errno (-r);
> -- 
> 2.31.1
> 

  reply	other threads:[~2022-11-29  8:56 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-29  3:16 [PATCH] linux: Change syscall return value to long int XingLi
2022-11-29  8:55 ` Szabolcs Nagy via Libc-alpha [this message]
2022-11-29  9:12   ` Xi Ruoyao via Libc-alpha
2022-11-29  9:37     ` Szabolcs Nagy via Libc-alpha
2022-11-29 10:54       ` XingLi

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

  List information: https://www.gnu.org/software/libc/involved.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Y4XI/A3whKF8BbA2@arm.com \
    --to=libc-alpha@sourceware.org \
    --cc=adhemerval.zanella@linaro.org \
    --cc=caiyinyu@loongson.cn \
    --cc=hejinyang@loongson.cn \
    --cc=i.swmail@xen0n.name \
    --cc=lixing@loongson.cn \
    --cc=szabolcs.nagy@arm.com \
    --cc=wanghongliang@loongson.cn \
    --cc=xry111@xry111.site \
    /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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).