Re: endless error loop

[Sam S. <sd...@gn...>]

> * Bruno Haible <oe...@py...t> [2004-05-28 19:21:05 +0200]:
>
> Sam wrote:
>> here is what I get with the Solaris/Sparcv9 64-bit executable you gave
>> me:
>>
>> Break 37 [38]>
>> *** - UNIX error 11 (EWOULDBLOCK): Operation would block
>
> Smells like either the LISTEN function or the NO-HANG things. Both end
> up using an ioctl().

I got it when I just started an interactive process:
$ clisp

> I built the binary on Solaris 9, you are on Solaris 8, right?

yes,
SunOS monty-burns 5.8 Generic_108528-24 sun4u sparc SUNW,Sun-Blade-1000

actually, a clisp-2.33-patched build is in progress right now!
(compiling compiler.lisp)

> In this in xterm, with stdin/stdout being a tty, or in Emacs, where
> stdout is a pipe?

this is an w2k/cygwin/bash in a w32 console window (not a cygwin xterm)
plus "rlogin monty".

> An strace of what the binary is doing w.r.t. stdin and stdout would be
> useful. (strace is called "truss" on Solaris.)

alas, I killed the window which exhibited this behavior, so I cannot
reproduce the bug anymore.  too bad.
I will take a mental note and will strace/truss when this happens again.

>> maybe it would be possible to limit the number of nested errors?
>
> It is already limited. At some point you'll get a "Lisp stack
> overflow: RESET".

Actually, I got a segfault instead - on about 150th error.


-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
When we break the law, they fine us, when we comply, they tax us.