Bug #7265 Test failure: 'func_time' (microseconds deleted in PS protocol)
Submitted: 14 Dec 2004 13:26 Modified: 14 Jan 2005 18:44
Reporter: Joerg Bruehe Email Updates:
Status: Closed Impact on me:
None 
Category:MySQL Server Severity:S3 (Non-critical)
Version:5.0.3-pre OS:Linux (Linux / Athlon)
Assigned to: Konstantin Osipov CPU Architecture:Any

[14 Dec 2004 13:26] Joerg Bruehe
Description:
Test 'func_time' fails for me in current 5.0,
but only if I run the suite with '--ps-protocol'.

This is a non-debug build: 'compile-pentium-max'.

Failure report:
-------------------------------------------------------
*** r/func_time.result  2004-12-10 18:05:58.000000000 +0300
--- r/func_time.reject  2004-12-14 12:50:02.000000000 +0300
***************
*** 553,559 ****
  2003-01-02 00:00:01
  select timestampadd(SQL_TSI_FRAC_SECOND, 1, date) from t1;
  timestampadd(SQL_TSI_FRAC_SECOND, 1, date)
! 2003-01-02 00:00:00.000001
  select timestampdiff(MONTH, '2001-02-01', '2001-05-01') as a;
  a
  3
--- 553,559 ----
  2003-01-02 00:00:01
  select timestampadd(SQL_TSI_FRAC_SECOND, 1, date) from t1;
  timestampadd(SQL_TSI_FRAC_SECOND, 1, date)
! 2003-01-02 00:00:00
  select timestampdiff(MONTH, '2001-02-01', '2001-05-01') as a;
  a
  3
-------------------------------------------------------

Latest changeset on my PC:
ChangeSet@1.1738, 2004-12-14 13:41:32+03:00, gluh@gluh.mysql.r18.ru
  Fix for bug #7223: information_schema: error in "views"

The error is older, as I got it already yesterday but did not yet report.

How to repeat:
Run the test suite with '--ps-protocol'.
[14 Jan 2005 18:44] Joerg Bruehe
I cannot reproduce it any more, seems to have been corrected (there were corrections for the '--ps-protocol' case).