MySQL Bugs: #21089: mysqltest doesn't print warnings produced by mysql

Bug #21089	mysqltest doesn't print warnings produced by mysql_prepare(stmt-with-resultset)
Submitted:	17 Jul 2006 1:48	Modified:	12 Jan 2011 12:05
Reporter:	Sergey Petrunya	Email Updates:
Status:	Duplicate	Impact on me:	None
Category:	Tools: MTR / mysql-test-run	Severity:	S3 (Non-critical)
Version:		OS:	Any
Assigned to:	Bjørn Munch	CPU Architecture:	Any

Description:
mysqltest client doesn't print warnings produced by mysql_prepare(statement-with-resultset).

How to repeat:
Create a .test file with this contents:
--- t/_a.test ---
create table t1 (a int);
insert into t1 values (0),(1),(2),(3),(4),(5),(6),(7),(8),(9);

select a+1 a from t1 group by a having a > 5;
--- t/_a.test ends---

run 
./mysql-test-run --record t/_a.test

Then run ./mysql-test-run --ps-protocol t/_a.test, and get this test failure:

-------------------------------------------------------
*** r/_c.result Mon Jul 17 08:13:34 2006
--- r/_c.reject Mon Jul 17 08:13:52 2006
***************
*** 8,11 ****
  10
  Warnings:
  Warning       1052    Column 'a' in group statement is ambiguous
- Warning       1052    Column 'a' in having clause is ambiguous
--- 8,10 ----
-------------------------------------------------------

The following fix makes the mysql_prepare() warnings to be printed into test output:

 ===== mysqltest.c 1.237 vs edited =====
*** /home/psergey/mysql-5.0-opt-check/BitKeeper/tmp/bk_mysqltest.c-1.237_CH72IE 2006-07-17 10:14:37 +04:00
--- edited/mysqltest.c  2006-07-17 09:46:36 +04:00
***************
*** 3985,3991 ****
        mysql_free_result(res);     /* Free normal result set with meta data */
  
        /* Clear prepare warnings */
!       dynstr_set(&ds_prepare_warnings, NULL);
      }
      else
      {
--- 3985,3991 ----
        mysql_free_result(res);     /* Free normal result set with meta data */
  
        /* Clear prepare warnings */
!       //psergey: don't: dynstr_set(&ds_prepare_warnings, NULL);
      }
      else
      {
--end of file--

However, we can't just push it. The problem is that certain warnings are retruned for both mysql_stmt_prepare() and mysql_stmt_execute(). I've already caught those:
 Incorrect date value: 'zzz' for column 'f1' at row 1
 Column 'c2' in group statement is ambiguous

Returning a warning in both mysql_stmt_prepare() and mysql_stmt_execute() causes this:
 - with --ps-protocol test result contains two copies warnings.
 - without --ps-protocol test result contains one warning.
while the testsuite (correctly) expects output to be the same with and without --ps-protocol.

 - without

Duplicate of recently fixed Bug #47902