public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Stian Halseth <stian@itx.no>
To: libc-alpha@sourceware.org
Cc: adhemerval.zanella@linaro.org, sam@gentoo.org,
	Stian Halseth <stian@itx.no>
Subject: [PATCH v2] Mark swapcontext returns_twice by default [BZ #23130]
Date: Fri,  4 Sep 2026 17:29:55 +0200	[thread overview]
Message-ID: <20260904152955.4082187-1-stian@itx.no> (raw)
In-Reply-To: <c959f881-f159-4855-9d11-e1c8c670db95@linaro.org>

swapcontext does not restore every callee-saved register on all targets.
On sparc the %l and %i registers are not in the saved context at all:
they live in the register window save area of the calling frame, which
another pass through that frame can overwrite.  stdlib/tst-setcontext2
fails on sparc64 for this reason at -O2 with PIC, where gcc keeps
&global in %i5 across swapcontext and reuses %i5 as the GOT base on
another path through the same frame.

Mark swapcontext returns_twice in the generic bits/indirect-return.h, so
every target that does not override it is covered.  x86 and aarch64 keep
their own headers and are unchanged.

string/tst-xbzero-opt.c calls swapcontext from an always_inline
function, which a returns_twice callee forbids; key its guard on the
availability of indirect_return rather than on __CET__.  It is the only
such site in the tree.

Tested on sparc64: full make check is clean, with no test that passed
before failing after.

Signed-off-by: Stian Halseth <stian@itx.no>
---
Changes in v2:
 - drop the sparc-specific rationale from the generic header
 - use returns_twice unconditionally rather than preferring
   indirect_return

I left the tst-xbzero-opt.c guard keyed on indirect_return.  The
attribute is x86/aarch64-only in practice -- __has_attribute reports 1
on x86_64 and 0 on sparc64 -- and both of those targets have their own
headers, so the guard still selects exactly the targets that end up with
returns_twice.

On sparc64 the generic header resolves to returns_twice either way, so
the make check result from v1 stands unchanged.

 bits/indirect-return.h  | 9 +++++++--
 string/tst-xbzero-opt.c | 7 +++----
 2 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/bits/indirect-return.h b/bits/indirect-return.h
index 819bd25b..ba2efb06 100644
--- a/bits/indirect-return.h
+++ b/bits/indirect-return.h
@@ -21,5 +21,10 @@
 #endif
 
 /* __INDIRECT_RETURN is used on swapcontext to indicate if it requires
-   special compiler treatment.  */
-#define __INDIRECT_RETURN
+   special compiler treatment.  swapcontext does not restore every
+   callee-saved register on all targets.  */
+#if __glibc_has_attribute (__returns_twice__)
+# define __INDIRECT_RETURN __attribute__ ((__returns_twice__))
+#else
+# define __INDIRECT_RETURN
+#endif
diff --git a/string/tst-xbzero-opt.c b/string/tst-xbzero-opt.c
index 3aa13d03..26e2ec1c 100644
--- a/string/tst-xbzero-opt.c
+++ b/string/tst-xbzero-opt.c
@@ -111,10 +111,9 @@ use_test_buffer (unsigned char *buf)
 /* Always check the test buffer immediately after filling it; this
    makes externally visible side effects depend on the buffer existing
    and having been filled in.  */
-#if defined __CET__ && !__glibc_has_attribute (__indirect_return__)
-/* Note: swapcontext returns via indirect branch when SHSTK is enabled.
-   Without indirect_return attribute, swapcontext is marked with
-   returns_twice attribute, which prevents always_inline to work.  */
+#if !__glibc_has_attribute (__indirect_return__)
+/* Without indirect_return, swapcontext is marked returns_twice, which
+   prevents always_inline from working.  */
 # define ALWAYS_INLINE
 #else
 # define ALWAYS_INLINE	__attribute__ ((always_inline))
-- 
2.43.0


  reply	other threads:[~2026-09-04 15:30 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-02 12:43 [PATCH] Mark swapcontext returns_twice without indirect_return " Stian Halseth
2026-09-04 14:20 ` Adhemerval Zanella Netto
2026-09-04 15:29   ` Stian Halseth [this message]
2026-09-04 18:32     ` [PATCH v2] Mark swapcontext returns_twice by default " Adhemerval Zanella Netto
2026-09-15 21:00       ` Stian Halseth
2026-09-16 12:02         ` Adhemerval Zanella Netto
2026-09-16 12:33           ` Stian Halseth
2026-09-21 12:33     ` Andreas Schwab
2026-09-21 13:05       ` Stian Halseth
2026-09-21 13:14         ` Andreas Schwab
2026-09-21 20:36           ` Stian Halseth
2026-09-21 20:39           ` [PATCH] string: Drop always_inline from tst-xbzero-opt " Stian Halseth
2026-09-25 10:43             ` Stian Halseth
2026-09-25 13:03               ` Adhemerval Zanella Netto
2026-09-25 11:11             ` Adhemerval Zanella Netto

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=20260904152955.4082187-1-stian@itx.no \
    --to=stian@itx.no \
    --cc=adhemerval.zanella@linaro.org \
    --cc=libc-alpha@sourceware.org \
    --cc=sam@gentoo.org \
    /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).