mailing list of musl libc
 help / color / mirror / code / Atom feed
From: Szabolcs Nagy <nsz@port70.net>
To: Alexander Monakov <amonakov@ispras.ru>
Cc: musl@lists.openwall.com
Subject: Re: [musl] [PATCH 1/3] setjmp: fix x86-64 longjmp argument adjustment
Date: Wed, 12 Aug 2020 23:44:05 +0200	[thread overview]
Message-ID: <20200812214405.GE879655@port70.net> (raw)
In-Reply-To: <alpine.LNX.2.20.13.2008121624170.7727@monopod.intra.ispras.ru>

[-- Attachment #1: Type: text/plain, Size: 734 bytes --]

* Alexander Monakov <amonakov@ispras.ru> [2020-08-12 16:29:59 +0300]:
> By the way, you should check if other 64-bit ports in musl exhibit
> the same issue. The AArch64 port definitely does, for example.
> 
> Here's a standalone testcase, needs -O2 so test_lj passes its second
> argument unchanged:

yes it seems wrong, see attached patch (with that the test passes).
thanks.

> 
> #include <setjmp.h>
> 
> static void test_lj(jmp_buf jb, long lv)
> {
>         longjmp(jb, lv);
> }
> 
> int main()
> {
>         void (*volatile ptest)(jmp_buf, long) = 0;
>         jmp_buf jb;
> 
>         int v = setjmp(jb);
>         ptest = ptest ? 0 : test_lj;
>         if (ptest) ptest(jb, 1l<<32);
>         return !v;
> }
> 
> 
> Alexander

[-- Attachment #2: 0001-aarch64-fix-setjmp-return-value.patch --]
[-- Type: text/x-diff, Size: 821 bytes --]

From c1b4a33e7cb1f4df0cc5f8237930ff81786cfbde Mon Sep 17 00:00:00 2001
From: Szabolcs Nagy <nsz@port70.net>
Date: Wed, 12 Aug 2020 21:00:26 +0000
Subject: [PATCH] aarch64: fix setjmp return value

longjmp should set the return value of setjmp, but 64bit
registers were used for the 0 check while the type is int.

use the code that gcc generates for return val ? val : 1;
---
 src/setjmp/aarch64/longjmp.s | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/src/setjmp/aarch64/longjmp.s b/src/setjmp/aarch64/longjmp.s
index 7c4655fa..b22042a2 100644
--- a/src/setjmp/aarch64/longjmp.s
+++ b/src/setjmp/aarch64/longjmp.s
@@ -18,7 +18,6 @@ longjmp:
 	ldp d12, d13, [x0,#144]
 	ldp d14, d15, [x0,#160]
 
-	mov x0, x1
-	cbnz x1, 1f
-	mov x0, #1
+	cmp w1, 0
+	csinc w0, w1, wzr, ne
 1:	br x30
-- 
2.28.0


  reply	other threads:[~2020-08-12 21:44 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-11 18:11 Alexander Monakov
2020-08-11 18:11 ` [musl] [PATCH 2/3] setjmp: avoid useless REX-prefix on xor %eax, %eax Alexander Monakov
2020-08-11 18:11 ` [musl] [PATCH 3/3] setjmp: optimize x86 longjmp epilogues Alexander Monakov
2020-08-11 18:45 ` [musl] [PATCH 1/3] setjmp: fix x86-64 longjmp argument adjustment Rich Felker
2020-08-12 13:29   ` Alexander Monakov
2020-08-12 21:44     ` Szabolcs Nagy [this message]
2020-08-12 22:13       ` Szabolcs Nagy
2020-08-13  8:01     ` Jens Gustedt
2020-08-13  8:53       ` Alexander Monakov
2020-08-11 23:10 ` [musl] [PATCH] setjmp: optimize x86-64 longjmp prologues Alexander Monakov
2020-08-12 11:34   ` [musl] [PATCH 4/3] setjmp: optimize " Alexander Monakov

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=20200812214405.GE879655@port70.net \
    --to=nsz@port70.net \
    --cc=amonakov@ispras.ru \
    --cc=musl@lists.openwall.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.
Code repositories for project(s) associated with this public inbox

	https://git.vuxu.org/mirror/musl/

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 NNTP newsgroup(s).