Re: [HACKERS] improper call to spi_printtup ???

2004-06-29 Thread Darko Prenosil
> - Original Message - > From: "Tom Lane" <[EMAIL PROTECTED]> > To: "Darko Prenosil" <[EMAIL PROTECTED]> > Cc: <[EMAIL PROTECTED]> > Sent: Monday, June 28, 2004 9:39 PM > Subject: Re: [HACKERS] improper call to spi_printtup ??? >

Re: [HACKERS] improper call to spi_printtup ???

2004-06-28 Thread Darko Prenosil
sprintf(values[12],"%c",inf->result->tupdesc->attrs[call_cntr]->attalign); sprintf(values[13],"%i",inf->result->tupdesc->attrs[call_cntr]->attnotnull); sprintf(values[14],"%i",inf->result->tupdesc->attrs[call_cntr]->atthasdef);

Re: [HACKERS] improper call to spi_printtup ???

2004-06-28 Thread Tom Lane
Darko Prenosil <[EMAIL PROTECTED]> writes: > Anyone knows what I'm doing wrong ? Well, when you didn't show us the text of the function, no. However, a reasonable bet would be that you used SPI inside the function and did not use it correctly, leaving the SPI state corrupted when control got back

[HACKERS] improper call to spi_printtup ???

2004-06-28 Thread Darko Prenosil
I have set returning function written in 'c', declared as: CREATE OR REPLACE FUNCTION check_view (text,text) RETURNS setof pg_attribute AS '/usr/local/pgsql/lib/libplpq.so','check_view' LANGUAGE 'c' WITH (isstrict); When I call this function from psql : SELECT attrelid,attnum FROM che