This is the mail archive of the glibc-bugs@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug nptl/22563] __cancel_jmp_buf in __pthread_unwind_buf_t is incompatible with setjmp/longmp


https://sourceware.org/bugzilla/show_bug.cgi?id=22563

Carlos O'Donell <carlos at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
                 CC|                            |carlos at redhat dot com
         Resolution|FIXED                       |---
   Target Milestone|2.27                        |---

--- Comment #25 from Carlos O'Donell <carlos at redhat dot com> ---
The following commits:
commit f81ddabffd76ac9dd600b02adbf3e1dac4bb10ec
commit cba595c350e52194e10c0006732e1991e3d0803b

Will be reverted within the next 24-hours since they cause an ABI regression.

Existing binaries have only 104-bytes of stack space in which to store
__pthread_unwind_buf_t, and as such cannot store the new sized object with the
larger 128-byte __sigset_t saved_mask.

The cleanest solution is to version __pthread_register_cancel so it can deal
with the smaller sized object (as if CET was not enabled). Likewise
__pthread_unwind_next which looks at priv.data.prev.

This will have to wait for glibc 2.28.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]