Re: Again: Bug with longs in Pythondriver

Andre Reitz wrote:
> under special conditions the pythondriver ends in a segmentation fault.
[...]
> Please look at 7. and 8. These querys result in a Segmentation fault,
> please note that 9. and 10. do not result in a segfault, but 8. does.
> (9 and 10 select the same rows a 8. does, but only 8. crashes)

Would it be possible to create a short python script that demonstrates 
this bug. The important details are probably the row size of the table, 
the individual sizes of the LONGs and the PACKET_SIZE kernel parameter.

Does this error occurs when the SELECT is executed, when the row is 
fetched or when the content of the LONG is read?

A kernel trace (http://sapdb.2scale.net/moin.cgi/VTrace) would be also 
helpful tyo make sure that the demo script behaves exactly the same.

I guess that the Python driver doesn't handle some border condition. 
Case 8. and 9. probably return the rows in a different order, so the 
layout inside the communication packet wil look different.

Daniel Dittmar

-- 
Daniel Dittmar
SAP DB, SAP Labs Berlin
[email protected]
http://www.sapdb.org