Incompatibility in Inline TLS Assembly on Alpine 3.22 with zend_jit_ir.c #18743

Activity

1. thecaliskan

   added

   Bug

   Status: Needs Triage

   on Jun 3, 2025
2. arnaud-lb commented on Jun 3, 2025

   

   arnaud-lb

   on Jun 3, 2025

   Member

   More actions

   Thank you for reporting this. This is the same root cause as #15074 (see #15074 (comment)), and will be fixed by #18660 (specifically ba04798).
3. ndossche commented on Jun 3, 2025

   

   ndossche

   on Jun 3, 2025

   Last edited by ndossche

   Member

   More actions

   @arnaud-lb But this can be fixed on lower branches, and I think we should.
   For this particular issue the following suffices (on PHP-8.3) (but would need to generalise to other platforms too ofc as we would need to change other code paths with the same assembly too):

   diff --git a/ext/opcache/jit/zend_jit_x86.dasc b/ext/opcache/jit/zend_jit_x86.dasc
   index 1f1abb59a1c..6ea8521ba6d 100644
   --- a/ext/opcache/jit/zend_jit_x86.dasc
   +++ b/ext/opcache/jit/zend_jit_x86.dasc
   @@ -2910,7 +2910,7 @@ static int zend_jit_setup(void)
    
    		__asm__(
    			"leaq _tsrm_ls_cache@tlsgd(%%rip), %0\n"
   -			: "=a" (ti));
   +			: "=D" (ti));
    		tsrm_tls_offset = ti[1];
    		tsrm_tls_index = ti[0] * 8;
    #elif defined(__FreeBSD__)
   
4. ndossche

   added

   Category: JIT

   Status: Verified

   Category: TSRM

   and removed

   Status: Needs Triage

   on Jun 3, 2025
5. arnaud-lb commented on Jun 5, 2025

   

   arnaud-lb

   on Jun 5, 2025

   Member

   More actions

   @nielsdos I agree, and it's great that there is a simpler fix, as ba04798 felt a bit risky for release branches.

   The change you propose looks good to me. We don't need to use the entire code sequence as in ba04798, because gas is only checking a single instruction, and ld will only check the entire sequence when building opcache statically, which is not allowed in release branches.

   Glibc seems unaffected as we use @gottpoff, for which gas doesn't check the destination register: https://github.com/bminor/binutils-gdb/blob/ec181e1710e37007a8d95c284609bfaa5868d086/gas/config/tc-i386.c#L6869

   Do you want to take care of this or shall I?
6. ndossche commented on Jun 5, 2025

   

   ndossche

   on Jun 5, 2025

   Member

   More actions

   Thanks for the link, I'm on it.
7. ndossche

   added a commit that references this issue on Jun 5, 2025

   Fix phpGH-18743: Incompatibility in Inline TLS Assembly on Alpine 3.22

   ...

   This issue will close once commit cda883b is merged into the 'master' branch.

   cda883b
8. ndossche

   linked a pull request that will close this issueFix GH-18743: Incompatibility in Inline TLS Assembly on Alpine 3.22 #18779on Jun 5, 2025
9. ndossche

   added 2 commits that reference this issue on Jun 9, 2025

   Fix GH-18743: Incompatibility in Inline TLS Assembly on Alpine 3.22

   ...

   This issue will close once commit b3c8afe is merged into the 'master' branch.

   b3c8afe

   Merge branch 'PHP-8.3' into PHP-8.4

   ...

   This issue will close once commit 7a0beb4 is merged into the 'master' branch.

   7a0beb4
10. ndossche

    closed this as completedin b3c8afeon Jun 9, 2025
11. ndossche

    added a commit that references this issue on Jun 9, 2025

    Merge branch 'PHP-8.4'

    ...

    This issue will close once commit 4c72203 is merged into the 'master' branch.

    4c72203
12. andares commented on Jul 1, 2025

    

    andares

    on Jul 1, 2025

    More actions

    Thanks for the link, I'm on it.谢谢你的链接，我正在处理。

    Hello, could you kindly clarify whether this bug fix is scheduled for inclusion in PHP 8.4.9, or if it was already resolved in the 8.4.8 release?
13. tianon commented on Jul 1, 2025

    

    tianon

    on Jul 1, 2025

    Contributor

    More actions

    If you click on b3c8afe (the commit which closed this), you can see where GitHub tells you tags it's in (and you can see 8.4.9RC1 in there, so it's the former).

    👍React with 👍1Reacted by Andares Cui
14. andares commented on Jul 1, 2025

    

    andares

    on Jul 1, 2025

    More actions

    If you click on b3c8afe (the commit which closed this), you can see where GitHub tells you tags it's in (and you can see 8.4.9RC1 in there, so it's the former).

    Thanks, I got it.