Re: [PATCHES] [HACKERS] 8.0.0beta5 FailedAssertion (Crash) when casting composite types
I have applied the attached patch to HEAD in order to fix the problems
discussed in this thread:
http://archives.postgresql.org/pgsql-hackers/2004-12/msg00187.php
regards, tom lane
Index: src/backend/executor/execQual.c
===
RCS file: /cvsroot/pgsql/src/backend/executor/execQual.c,v
retrieving revision 1.169
diff -c -r1.169 execQual.c
*** src/backend/executor/execQual.c 22 Sep 2004 17:41:50 - 1.169
--- src/backend/executor/execQual.c 11 Dec 2004 16:26:02 -
***
*** 87,92
--- 87,95
bool *isNull, ExprDoneCond *isDone);
static Datum ExecEvalAnd(BoolExprState *andExpr, ExprContext *econtext,
bool *isNull, ExprDoneCond *isDone);
+ static Datum ExecEvalConvertRowtype(ConvertRowtypeExprState *cstate,
+
ExprContext *econtext,
+ bool
*isNull, ExprDoneCond *isDone);
static Datum ExecEvalCase(CaseExprState *caseExpr, ExprContext *econtext,
bool *isNull, ExprDoneCond *isDone);
static Datum ExecEvalCaseTestExpr(ExprState *exprstate,
***
*** 428,434
*
*Returns a Datum whose value is the value of a range
*variable with respect to given expression context.
! * */
static Datum
ExecEvalVar(ExprState *exprstate, ExprContext *econtext,
bool *isNull, ExprDoneCond *isDone)
--- 431,438
*
*Returns a Datum whose value is the value of a range
*variable with respect to given expression context.
! *
! */
static Datum
ExecEvalVar(ExprState *exprstate, ExprContext *econtext,
bool *isNull, ExprDoneCond *isDone)
***
*** 1844,1849
--- 1848,1922
return BoolGetDatum(!AnyNull);
}
+ /*
+ *ExecEvalConvertRowtype
+ *
+ *Evaluate a rowtype coercion operation. This may require
+ *rearranging field positions.
+ *
+ */
+ static Datum
+ ExecEvalConvertRowtype(ConvertRowtypeExprState *cstate,
+ ExprContext *econtext,
+ bool *isNull, ExprDoneCond *isDone)
+ {
+ HeapTuple result;
+ Datum tupDatum;
+ HeapTupleHeader tuple;
+ HeapTupleData tmptup;
+ AttrNumber *attrMap = cstate->attrMap;
+ Datum *invalues = cstate->invalues;
+ char *innulls = cstate->innulls;
+ Datum *outvalues = cstate->outvalues;
+ char *outnulls = cstate->outnulls;
+ int i;
+ int outnatts = cstate->outdesc->natts;
+
+ tupDatum = ExecEvalExpr(cstate->arg, econtext, isNull, isDone);
+
+ /* this test covers the isDone exception too: */
+ if (*isNull)
+ return tupDatum;
+
+ tuple = DatumGetHeapTupleHeader(tupDatum);
+
+ Assert(HeapTupleHeaderGetTypeId(tuple) == cstate->indesc->tdtypeid);
+ Assert(HeapTupleHeaderGetTypMod(tuple) == cstate->indesc->tdtypmod);
+
+ /*
+* heap_deformtuple needs a HeapTuple not a bare HeapTupleHeader.
+*/
+ tmptup.t_len = HeapTupleHeaderGetDatumLength(tuple);
+ tmptup.t_data = tuple;
+
+ /*
+* Extract all the values of the old tuple, offsetting the arrays
+* so that invalues[0] is NULL and invalues[1] is the first
+* source attribute; this exactly matches the numbering convention
+* in attrMap.
+*/
+ heap_deformtuple(&tmptup, cstate->indesc, invalues + 1, innulls + 1);
+ invalues[0] = (Datum) 0;
+ innulls[0] = 'n';
+
+ /*
+* Transpose into proper fields of the new tuple.
+*/
+ for (i = 0; i < outnatts; i++)
+ {
+ int j = attrMap[i];
+
+ outvalues[i] = invalues[j];
+ outnulls[i] = innulls[j];
+ }
+
+ /*
+* Now form the new tuple.
+*/
+ result = heap_formtuple(cstate->outdesc, outvalues, outnulls);
+
+ return HeapTupleGetDatum(result);
+ }
/*
*ExecEvalCase
***
*** 2969,2974
--- 3042,3109
state = (ExprState *) gstate;
}
break;
+ case T_ConvertRowtypeExpr:
+ {
+ ConvertRowtypeExpr *convert =
Re: [PATCHES] Update french po files
Guillaume LELARGE wrote: > Here are french .po files. They have been proofreaded. > > Please, apply :) Done. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
Re: [PATCHES] PL/Perl document patch
David Fetter wrote: > Please find enclosed a patch to the pl/perl documents that clafies > the scope of global data and gives an example of storing a code > reference. Patch installed. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [PATCHES] [HACKERS] regression script/makefile exit failure
Andrew Dunstan wrote:
Peter Eisentraut wrote:
I seem to recall that there is a bug in the FreeBSD shell with line
breaks in traps. Try changing the above to
trap 'st=$?; echo status = $st; exit $st' 0
Thankyou Peter! I'd never have guessed something so obscure!
I have confirmed that this is the problem, and ascertained that the
Cygwin shell exhibits the same behaviour.
I will submit a patch for pg_regress.sh shortly (after a little
testing using buildfarm).
Attached patch has been tested on Cygwin and found to work as expected
when initdb fails during make check - I don't have access to a FreeBSD
machine to do a test buildfarm run.
cheers
andrew
Index: src/test/regress/pg_regress.sh
===
RCS file: /home/cvsmirror/pgsql/src/test/regress/pg_regress.sh,v
retrieving revision 1.50
diff -c -r1.50 pg_regress.sh
*** src/test/regress/pg_regress.sh 17 Nov 2004 18:06:04 - 1.50
--- src/test/regress/pg_regress.sh 11 Dec 2004 21:27:40 -
***
*** 238,255
# with the result of the last shell command before the `exit'. Hence
# we have to write `(exit x); exit' below this point.
! trap '
! savestatus=$?
if [ -n "$postmaster_pid" ]; then
kill -2 "$postmaster_pid"
wait "$postmaster_pid"
unset postmaster_pid
fi
rm -f "$TMPFILE" && exit $savestatus
! ' 0
! trap '
! savestatus=$?
echo; echo "caught signal"
if [ -n "$postmaster_pid" ]; then
echo "signalling fast shutdown to postmaster with pid $postmaster_pid"
--- 238,257
# with the result of the last shell command before the `exit'. Hence
# we have to write `(exit x); exit' below this point.
! exit_trap(){
! savestatus=$1
if [ -n "$postmaster_pid" ]; then
kill -2 "$postmaster_pid"
wait "$postmaster_pid"
unset postmaster_pid
fi
rm -f "$TMPFILE" && exit $savestatus
! }
! trap 'exit_trap $?' 0
!
! sig_trap() {
! savestatus=$1
echo; echo "caught signal"
if [ -n "$postmaster_pid" ]; then
echo "signalling fast shutdown to postmaster with pid $postmaster_pid"
***
*** 258,264
unset postmaster_pid
fi
(exit $savestatus); exit
! ' 1 2 13 15
--- 260,268
unset postmaster_pid
fi
(exit $savestatus); exit
! }
!
! trap 'sig_trap $?' 1 2 13 15
---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
Re: [PATCHES] Updated Turkish translations for PostgreSQL 8.0
Devrim GUNDUZ wrote: > We've updated the Turkish translations: Installed. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [PATCHES] Translation updates: 8.0: pg_dump-ru, initdb-ru, psql-ru; 7.4/8.0: postgres-ru
Serguei Mokhov wrote: > Attached please find newer Russian translation updates. Installed. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [PATCHES] Translation updates: pt_BR
Euler Taveira de Oliveira wrote: > This is a fully reviewed version of pt_BR message files. Please apply > it to HEAD. Done. > This another one, just resolved some fuzzy translations in 7.4. > Please apply it to 7.4 branch. Done. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 8: explain analyze is your friend
Re: [PATCHES] 8.0 RC1 NLS translate fo zh_CN
Weiping wrote: > Here is the translated string for zh_CN locale. Installed. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [PATCHES] Assorted spanish translations
Alvaro Herrera wrote: > I attach a tarball with the 10 message catalogs fully translated to > spanish. Installed. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 8: explain analyze is your friend
[PATCHES] Updated Turkish translations for PostgreSQL 8.0
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, We've updated the Turkish translations: http://postgresql.gunduz.org/translation/PostgreSQL-8.0/PgSQL-DilDosyalari/initdb-tr.po http://postgresql.gunduz.org/translation/PostgreSQL-8.0/PgSQL-DilDosyalari/pg_config-tr.po http://postgresql.gunduz.org/translation/PostgreSQL-8.0/PgSQL-DilDosyalari/pg_ctl-tr.po http://postgresql.gunduz.org/translation/PostgreSQL-8.0/PgSQL-DilDosyalari/pg_dump-tr.po http://postgresql.gunduz.org/translation/PostgreSQL-8.0/PgSQL-DilDosyalari/postgres-tr.po http://postgresql.gunduz.org/translation/PostgreSQL-8.0/PgSQL-DilDosyalari/psql-tr.po Could you please commit them? Regards, - -- Devrim GUNDUZ devrim~gunduz.orgdevrim.gunduz~linux.org.tr http://www.tdmsoft.com http://www.gunduz.org -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.1 (GNU/Linux) iD8DBQFBuvMBtl86P3SPfQ4RAhmKAKDlIyu9euRtjPlfjqPZVr2T7N+eZQCg0YLZ omNH7GJWQ/5I5MHUMkZPa+0= =IgmP -END PGP SIGNATURE- ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
