rpms/mysql8.4
5
0
Fork 0
Code Issues Pull Requests Releases Activity

Sigh ... this code is even more fragile than it appears. Try setting

Browse Source 
STACK_MIN_SIZE to 24K, not either 16K nor 32K, each of which fail for
    different reasons.
This commit is contained in:
Tom Lane 2008-01-08 05:41:31 +00:00 committed by Michal Schorm
parent bd5697ac43
commit 533c5cc42c
2 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
mysql-stack.patch
View File

@ -2,7 +2,9 @@ Talk about silly byte-shaving ... the "stock" value does not work on s390x,
and probably not on some other platforms either. Knock it up to twice the
pre-5.0.33 value to have some margin for future changes in compiler, glibc,
etc. Suspect this number if "execution_constants" regression test fails.
As of 5.0.45 and F9, it seems 16384 isn't enough either.
As of 5.0.45 and F9, it seems 16384 isn't enough anymore (ppc fails);
while 32768 is too much as it makes other regression tests fail.
diff -Naur mysql-5.0.38.orig/sql/mysql_priv.h mysql-5.0.38/sql/mysql_priv.h
@ -13,7 +15,7 @@ diff -Naur mysql-5.0.38.orig/sql/mysql_priv.h mysql-5.0.38/sql/mysql_priv.h
"execution_constants" test to pass.
*/
-#define STACK_MIN_SIZE 12000 // Abort if less stack during eval.
+#define STACK_MIN_SIZE 32768 // Abort if less stack during eval.
+#define STACK_MIN_SIZE 24576 // Abort if less stack during eval.
#define STACK_MIN_SIZE_FOR_OPEN 1024*80
#define STACK_BUFF_ALLOC 256 // For stack overrun checks

2
mysql.spec
View File

@ -482,7 +482,7 @@ fi
%{_mandir}/man1/mysql_client_test.1*
%changelog
* Mon Jan 7 2008 Tom Lane <tgl@redhat.com> 5.0.45-7
* Tue Jan 8 2008 Tom Lane <tgl@redhat.com> 5.0.45-7
- Unbelievable ... upstream still thinks that it's a good idea to have a
regression test that is guaranteed to begin failing come January 1.
- ... and it seems we need to raise STACK_MIN_SIZE again too.