MySQL Bugs: #28754: RPM builds differ from tar.gz in "ALTER ... RENAME" on views

Bug #28754	RPM builds differ from tar.gz in "ALTER ... RENAME" on views
Submitted:	29 May 2007 16:20	Modified:	14 Jun 2007 19:39
Reporter:	Joerg Bruehe	Email Updates:
Status:	Closed	Impact on me:	None
Category:	MySQL Server	Severity:	S2 (Serious)
Version:	5.1.19-beta, 5.0-bk	OS:	Linux (RPM only)
Assigned to:	Georgi Kodinov	CPU Architecture:	Any

Description:
Problem is new in 5.1.19-beta build, not observed before:

Test "rename" fails, in all 3 RPM builds and only in these, with this diff:

=====
Errors are (from /PATH/mysqltest-time) :
mysqltest: At line 85: query 'alter table v1 rename to v2' failed: 1347: 'test.v1' is not BASE TABLE
(the last lines may be the most important ones)
Result from queries before failure can be found in /PATH/mysql-test/var/log/rename.log
=====

In effect, this means bug#14959 is arisen again, but only in RPM builds.

How to repeat:
Build a RPM package of 5.1.19

Sorry - important detail missing:

This happens *only* in "embedded" build + test.

Could this be a specific effect of the RPM build environment, running the test ?

I've downloaded a fresh 5.1.19 RPM from the web site. I was able to build the binary RPMs on my Fedora 7 without any problem.

I was able to repeat the failure using a normal bk build and the following options : 
./mysql-test-run.pl  --comment=embedded --force --timer --embedded-server --skip-rpl --skip-ndbcluster t/rename.test

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/28392

ChangeSet@1.2523, 2007-06-08 13:42:10+03:00, gkodinov@magare.gmz +1 -0
  Bug #28754: RPM builds differ from tar.gz in "ALTER ... RENAME" on views
   When constructing the path to the original .frm file ALTER .. RENAME
   was unnecessary (and incorrectly) lowercasing the whole path when not 
   on a case-insensitive filesystem.
   This path should not be modified because of lower_case_table_names
   as it is already in the correct case according to that setting.
   Fixed by removing the lowercasing.
   Unfortunately testing this would require running the tests on a case
   sensitive filesystem in a directory that has uppercase letters.
   This cannot be guaranteed in all setups so no test case added.

The same problem exists in  5.0, so the fix is added to 5.0 and up.

Pushed into 5.1.20-beta

Pushed into 5.0.44

Noted in 5.0.44, 5.1.20 changelogs.