mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Ammar Faizi <ammarfaizi2@openresty.com>
To: "Thomas Weißschuh" <linux@weissschuh.net>
Cc: Willy Tarreau <w@1wt.eu>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Linux Kselftest Mailing List <linux-kselftest@vger.kernel.org>,
	LLVM Mailing List <llvm@lists.linux.dev>,
	Yichun Zhang <yichun@openresty.com>,
	Alviro Iskandar Setiawan <alviro.iskandar@gnuweeb.org>,
	Shuah Khan <shuah@kernel.org>,
	Nathan Chancellor <nathan@kernel.org>,
	Nick Desaulniers <nick.desaulniers+lkml@gmail.com>,
	Bill Wendling <morbo@google.com>,
	Justin Stitt <justinstitt@google.com>,
	gwml@gnuweeb.org
Subject: Re: [PATCH 3/4] selftests/nolibc: add abs() range test
Date: Mon, 27 Jul 2026 08:32:58 +0700	[thread overview]
Message-ID: <6a5f7481-124f-48ad-ab44-577b12d33592@openresty.com> (raw)
In-Reply-To: <8bdacaf3-7f93-4095-8875-4c13a273fe3d@t-8ch.de>

On 7/27/26 3:00 AM, Thomas Weißschuh wrote:
>> +/* abs(), labs() and llabs() over the whole range of their argument type */
>> +int test_abs_range(void)
>> +{
>> +	int i, ri;
>> +	long l, rl;
>> +	long long ll, rll;
> 
> Reverse xmas?

Fixed for the next revision.

>> +
>> +	/*
>> +	 * Both the inputs and the results have to stay opaque: the compiler
>> +	 * knows abs() and friends never return a negative value and would
>> +	 * otherwise fold the comparisons below at build time, which would also
>> +	 * hide the undefined behavior that is being tested for.
>> +	 */
>> +	i = INT_MIN; l = LONG_MIN; ll = LLONG_MIN;
>> +	__asm__ ("" : "+r" (i), "+r" (l), "+r" (ll));
> 
> We have _NOLIBC_OPTIMIZER_HIDE_VAR() for this.

Oh, I missed it.

> Maybe it can be made variadic.
Sounds good to me.

Does the following patch look good?

From: Ammar Faizi <ammarfaizi2@openresty.com>
Date: Sun, 26 Jul 2026 13:17:14 -0700
Subject: [PATCH] tools/nolibc: make _NOLIBC_OPTIMIZER_HIDE_VAR() variadic
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The macro only takes a single variable, so code that needs to hide
several at once has to either repeat it or open-code an asm statement
with a list of operands. Accept up to four variables and build the
operand list from them.

Existing single-argument users are unaffected; the code generated for
the callers in stdio.h is unchanged on i386 and x86-64.

Suggested-by: Thomas Weißschuh <linux@weissschuh.net>
Signed-off-by: Ammar Faizi <ammarfaizi2@openresty.com>
---
  tools/include/nolibc/compiler.h | 15 +++++++++++++--
  1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/tools/include/nolibc/compiler.h b/tools/include/nolibc/compiler.h
index f2d7a81d0d7c..ed69f6d26f56 100644
--- a/tools/include/nolibc/compiler.h
+++ b/tools/include/nolibc/compiler.h
@@ -77,8 +77,19 @@
  #  define __nolibc_static_assert(_t)
  #endif
  
-/* Make the optimizer believe the variable can be manipulated arbitrarily. */
-#define _NOLIBC_OPTIMIZER_HIDE_VAR(var)	__asm__ ("" : "+r" (var))
+#define __nolibc_hide1(_1)		"+r" (_1)
+#define __nolibc_hide2(_1, ...)		"+r" (_1), __nolibc_hide1(__VA_ARGS__)
+#define __nolibc_hide3(_1, ...)		"+r" (_1), __nolibc_hide2(__VA_ARGS__)
+#define __nolibc_hide4(_1, ...)		"+r" (_1), __nolibc_hide3(__VA_ARGS__)
+#define ___nolibc_hide_narg(_1, _2, _3, _4, N, ...) N
+#define __nolibc_hide_narg(...)		___nolibc_hide_narg(__VA_ARGS__, 4, 3, 2, 1)
+#define __nolibc_hide(N, ...)		__nolibc_hide##N(__VA_ARGS__)
+#define __nolibc_hide_n(N, ...)		__nolibc_hide(N, __VA_ARGS__)
+
+/* Make the optimizer believe the variables can be manipulated arbitrarily. */
+#define _NOLIBC_OPTIMIZER_HIDE_VAR(...)					\
+	__asm__ ("" : __nolibc_hide_n(__nolibc_hide_narg(__VA_ARGS__),	\
+				      __VA_ARGS__))
  
  #if __nolibc_has_feature(undefined_behavior_sanitizer)
  #  if defined(__clang__)

-- 
Ammar Faizi


  reply	other threads:[~2026-07-27  1:33 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-26 10:13 [PATCH 0/4] nolibc: syscall() and abs() fixes, plus a cleanup Ammar Faizi
2026-07-26 10:13 ` [PATCH 1/4] tools/nolibc: evaluate syscall() arguments before the arch macros Ammar Faizi
2026-07-26 20:16   ` Thomas Weißschuh
2026-07-27  1:21     ` Ammar Faizi
2026-07-27  3:42       ` Willy Tarreau
2026-07-28  6:53         ` Ammar Faizi
2026-07-27  3:30     ` Willy Tarreau
2026-07-27 16:18       ` Thomas Weißschuh
2026-07-28  6:51         ` Ammar Faizi
2026-07-26 10:13 ` [PATCH 2/4] tools/nolibc: stdlib: avoid signed overflow in abs() and friends Ammar Faizi
2026-07-26 14:13   ` David Laight
2026-07-26 16:01     ` Willy Tarreau
2026-07-27  7:32       ` David Laight
2026-07-27 16:15         ` Thomas Weißschuh
2026-07-28  8:46   ` David Laight
2026-07-28 16:31     ` Thomas Weißschuh
2026-07-28 16:37       ` Ammar Faizi
2026-07-26 10:13 ` [PATCH 3/4] selftests/nolibc: add abs() range test Ammar Faizi
2026-07-26 20:00   ` Thomas Weißschuh
2026-07-27  1:32     ` Ammar Faizi [this message]
2026-07-27  2:01       ` Ammar Faizi
2026-07-27 16:13         ` Thomas Weißschuh
2026-07-28  6:14           ` Ammar Faizi
2026-07-26 10:13 ` [PATCH 4/4] tools/nolibc: remove dead __ARCH_WANT_SYS_OLD_SELECT Ammar Faizi
2026-07-26 20:01   ` Thomas Weißschuh

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=6a5f7481-124f-48ad-ab44-577b12d33592@openresty.com \
    --to=ammarfaizi2@openresty.com \
    --cc=alviro.iskandar@gnuweeb.org \
    --cc=gwml@gnuweeb.org \
    --cc=justinstitt@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=linux@weissschuh.net \
    --cc=llvm@lists.linux.dev \
    --cc=morbo@google.com \
    --cc=nathan@kernel.org \
    --cc=nick.desaulniers+lkml@gmail.com \
    --cc=shuah@kernel.org \
    --cc=w@1wt.eu \
    --cc=yichun@openresty.com \
    /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

all inboxes | Powered by JetHome®