I copeid myself on this forward and never received it so
I am sending it again....

It has been over a week and I have not seen this posted.
I was just wondering if this got lost or something...

Thanks for your help!
Lynn

----- Forwarded message from Lynn.Tilby (AT) asu (DOT) edu -----
Date: Thu, 22 Jan 2004 21:56:57 -0700 (MST)
From: Lynn.Tilby (AT) asu (DOT) edu
Reply-To: Lynn.Tilby (AT) asu (DOT) edu
Subject: Lots of nan's
To: pgsql-novice (AT) postgresql (DOT) org

I am getting nan's all over the place...

I am running RH9 with the 2.4.20-27.9 #1 kernel, and Postgres 7.3

I have a debug macro as follows:

#define DBG_SQL(strng) if(DEBUG > 0) \
{ \
fprintf(DBG_OUT,"\n******DBG_SQL: %s\n",strng); \
fprintf(DBG_OUT,"sqlcaid = %s\n", sqlca.sqlcaid); \
fprintf(DBG_OUT,"sqlabc = %f\n", sqlca.sqlabc); \
fprintf(DBG_OUT,"sqlcode = %f\n", sqlca.sqlcode); \
fprintf(DBG_OUT,"sqlca.sqlerrm.sqlerrml = %d\n", sqlca.sqlerrm.sqlerrml); \
fprintf(DBG_OUT,"sqlca.sqlerrm.sqlerrmc = %s\n", sqlca.sqlerrm.sqlerrmc); \
fprintf(DBG_OUT,"sqlerrp = %s\n", sqlca.sqlerrp); \
fprintf(DBG_OUT,"sqlerrd = %f\n", sqlca.sqlerrd); \
fprintf(DBG_OUT,"sqlext = %s\n", sqlca.sqlext); \
} \

I get the following output when turning on debugging, not all of the time, but
only occasionally:

******DBG_SQL: after CONNECT TO pma
sqlcaid = SQLCA
sqlabc = nan
sqlcode = nan
sqlca.sqlerrm.sqlerrml = 0
sqlca.sqlerrm.sqlerrmc =
sqlerrp = NOT SET
sqlerrd = nan
sqlext =

******DBG_SQL: after WHENEVER sqlerror sqlprint
sqlcaid = SQLCA
sqlabc = nan
sqlcode = nan
sqlca.sqlerrm.sqlerrml = 0
sqlca.sqlerrm.sqlerrmc =
sqlerrp = NOT SET
sqlerrd = nan
sqlext =

******DBG_SQL: after CREATE TABLE stck_rnkg_tmp
sqlcaid = SQLCA
sqlabc = nan
sqlcode = nan
sqlca.sqlerrm.sqlerrml = 0
sqlca.sqlerrm.sqlerrmc =
sqlerrp = NOT SET
sqlerrd = nan
sqlext =
..
..
..
..

PLEASE NOTE! I have searched for the string 'nan' in all tables, data and code
and have not been able to find it. I have also searched the .c output
from ecpg and not found anything. Nan stands for not a number.

Further down in the code I get the following error:

sql error 'ERROR: Attribute "nan" not found' in line 1235.

Line 1235 in the code is simply C code not SQL, and is C code for 20+ lines
both above and below line 1235. I have also looked at the .c output from
ecpg but have not found anything remotely connected.

Further down in the code when printing out further debugging info I get this:

tot_sl_prce = nan
t_expr_c_date_num = 20031017
t_by_date_num = 731490
tc_exp_date_num = 20031017
tc_exp_date_dys = 731506
t_sl_comm = 0.000000
tot_crnt_prce = 0.000000
tot_sl_prce = nan
tot_by_prce = nan
t_prft_los = nan
t_pct_prft_los = nan

Still more variables take on the value of nan. Then a little further I get
this:

sqlcaid = SQLCA
sqlabc = 0.000000
sqlcode = 0.000000
sqlca.sqlerrm.sqlerrml = 49
sqlca.sqlerrm.sqlerrmc = 'ERROR: Attribute "nan" not found' in line 1235.
sqlerrp = NOT SET
sqlerrd = 0.000000
sqlext =



In doing searches on the Pgsql site I found that 'attribute' is a historical
term for column heading; but since I don't have any columns of the name nan
I can't see where 'nan' cannot be found. I found nothing on the pgsql site
for nan, but, In searching google for nan+postgres I got about 1180 hits; but,
none of the ones I looked at were current so I don't know if this problem has
been fixed yet.

The really interesting thing is when I let this prog run it consistently
crashes RedHat and I have to do a hardware reset to reboot.

How do I solve this problem?

Thanks for your help!!!!
Lynn


----- End forwarded message -----

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org

On Mon, 9 Feb 2004 Lynn.Tilby (AT) asu (DOT) edu wrote:

The sqlca.h file I see seems to imply that sqlabc and sqlcode
are longs and sqlerrd is long[6]. If that's right, writing them
with %f is a mistake (technically undefined behavior I think)

---------------------------(end of broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to majordomo (AT) postgresql (DOT) org