Linux s390 Architecture development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "André Almeida" <andrealmeid@igalia.com>
Cc: linux-s390@vger.kernel.org, "Vasily Gorbik" <gor@linux.ibm.com>,
	"Heiko Carstens" <hca@linux.ibm.com>,
	"Alexander Gordeev" <agordeev@linux.ibm.com>,
	"Christian Borntraeger" <borntraeger@linux.ibm.vom>
Subject: Re: [PATCH v5 04/13] arm: Use the common syscall table
Date: Wed, 12 Aug 2026 18:10:03 +0000	[thread overview]
Message-ID: <20260812181003.D59DB1F000E9@smtp.kernel.org> (raw)
In-Reply-To: <20260812-tonyk-syscall_table-v5-4-3d383927c609@igalia.com>

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Using the common syscall table on 32-bit ARM breaks OABI compatibility for several time64 and wait syscalls due to missing or weak-linked `compat_sys_*` entry points.
- [High] Process substitution `< <(...)` breaks POSIX `/bin/sh` compatibility, causing kernel build failures on systems using dash.
--

--- Patch [4]: [PATCH v5 04/13] arm: Use the common syscall table ---
commit 4446b58b5d8e69028beb1ab8dad93b78f955eb73
Author: André Almeida <andrealmeid@igalia.com>

arm: Use the common syscall table

This commit updates the ARM architecture to use the common syscall table.
It replaces duplicated ARM-specific syscalls with the common ones by pulling
in syscall_common.tbl and updating the ABI list.

> diff --git a/arch/arm/tools/Makefile b/arch/arm/tools/Makefile
> index 28b6da8ac5f64..fd73bf6db3bfe 100644
> --- a/arch/arm/tools/Makefile
> +++ b/arch/arm/tools/Makefile
[ ... ]
> @@ -38,11 +40,13 @@ $(kapi)/mach-types.h: $(src)/gen-mach-types $(src)/mach-types FORCE
>  	$(call if_changed,gen_mach)
>  
>  quiet_cmd_syshdr = SYSHDR  $@
> -      cmd_syshdr = $(CONFIG_SHELL) $(syshdr) --abis common,$* \
> +      cmd_syshdr = $(CONFIG_SHELL) $(syshdr) --abis $(abis),$* \
> +		   --common-tbl $(syscalltbl_common) \
>  		   --offset __NR_SYSCALL_BASE $< $@

[Severity: High]
Does this execute scripts containing non-POSIX bashisms?

The scripts/syscallhdr.sh script uses process substitution which requires bash:

scripts/syscallhdr.sh:gen_hdr() {
    ...
    done < <(grep -E "^[0-9A-Fa-fXx]+[[:space:]]+$abis" "$input")
}

scripts/syscalltbl.sh also uses a similar construct in gen_tbl().

This breaks POSIX /bin/sh compatibility and can cause kernel build failures
reporting "Syntax error: redirection unexpected" on distributions that use
dash as their default shell.

>  quiet_cmd_systbl = SYSTBL  $@
> -      cmd_systbl = $(CONFIG_SHELL) $(systbl) --abis common,$* $< $@
> +      cmd_systbl = $(CONFIG_SHELL) $(systbl) --abis $(abis),$* \
> +		   --common-tbl $(syscalltbl_common) $< $@

[Severity: High]
Does pulling in the common syscall table break OABI compatibility for
32-bit ARM?

The common table populates the compat column with compat_sys_* functions
intended for CONFIG_COMPAT, for example:

scripts/syscall_common.tbl:
    413  32  pselect6_time64  sys_pselect6  compat_sys_pselect6_time64

However, ARM OABI redefines __SYSCALL_WITH_COMPAT to use the compat column
specifically for its own OABI wrappers, not for 64-bit-on-32-bit compat
wrappers:

arch/arm/kernel/entry-common.S:sys_oabi_call_table() {
    #undef __SYSCALL_WITH_COMPAT
    #define __SYSCALL_WITH_COMPAT(nr, native, compat) __SYSCALL(nr, compat)
}

This clash causes build failures like "undefined reference to
compat_sys_pselect6_time64" during linking when CONFIG_OABI_COMPAT is enabled.
For syscalls with weak stubs in kernel/sys_ni.c (like io_pgetevents_time64),
the build succeeds but OABI binaries will silently receive -ENOSYS at runtime.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260812-tonyk-syscall_table-v5-0-3d383927c609@igalia.com?part=4

  reply	other threads:[~2026-08-12 18:10 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-12 17:38 [PATCH v5 00/13] syscalls: Add a shared table for all archs André Almeida
2026-08-12 17:38 ` [PATCH v5 01/13] syscalls: Make --abis parsing more robust André Almeida
2026-08-12 17:50   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 02/13] syscalls: Create unified partial table for all archs André Almeida
2026-08-12 17:49   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 03/13] scripts/syscall.tbl: Use the common table André Almeida
2026-08-12 17:53   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 04/13] arm: Use the common syscall table André Almeida
2026-08-12 18:10   ` sashiko-bot [this message]
2026-08-12 17:38 ` [PATCH v5 05/13] s390: " André Almeida
2026-08-12 17:50   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 06/13] sparc: " André Almeida
2026-08-12 17:48   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 07/13] mips: Remove duplicated syscallnr.sh André Almeida
2026-08-12 17:48   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 08/13] mips: Get rid of custom mips ABIs for syscall tables André Almeida
2026-08-12 17:49   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 09/13] mips: Use the common syscall table André Almeida
2026-08-12 17:53   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 10/13] syscalls: Add an option for offsetting the common table André Almeida
2026-08-12 17:56   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 11/13] alpha: Define entry point for set_mempolicy_home_node syscall André Almeida
2026-08-12 17:57   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 12/13] alpha: Remove alpha_ prefix from custom syscall entries André Almeida
2026-08-12 17:58   ` sashiko-bot
2026-08-12 17:38 ` [PATCH v5 13/13] alpha: Use the common syscall table André Almeida
2026-08-12 18:01   ` sashiko-bot

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=20260812181003.D59DB1F000E9@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=agordeev@linux.ibm.com \
    --cc=andrealmeid@igalia.com \
    --cc=borntraeger@linux.ibm.vom \
    --cc=gor@linux.ibm.com \
    --cc=hca@linux.ibm.com \
    --cc=linux-s390@vger.kernel.org \
    --cc=sashiko-reviews@lists.linux.dev \
    /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