Ilya Zakharevich on Mon, 21 Jul 2003 05:04:26 -0700


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: ^C problem


On Fri, Jul 18, 2003 at 08:28:13PM -0400, Igor Schein wrote:
> Hi,
> 
> once in a while I get into situation when I am running a loop in gp,
> then ^C out of it, then change the loop command, then run again etc.,
> and all of a sudden ^C stops working.  trap() is not involved at all,

I think it may be due to difference between semantic of BSD and SysV
signals.  A similar situation existed on OS/2 until I added
-Zsysv-signals to the compiler flags; with this flag the signals work
better, but still not 100% reliably.

So it looks like the PARI source (mostly?) assumes one particulare
semantic (SysV).  And, as usual in such situations, it may have bugs
related to the difference of semantic - some pieces of code may be
written by people expecting one semantic, another one by the others.

Hope this helps,
Ilya