gcl-devel
[Top][All Lists]
Advanced

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

[Gcl-devel] Re: OpenBSD: huge allocation?


From: Magnus Henoch
Subject: [Gcl-devel] Re: OpenBSD: huge allocation?
Date: Sun, 28 Mar 2004 16:35:21 +0200
User-agent: Gnus/5.1006 (Gnus v5.10.6) Emacs/21.3 (gnu/linux)

Camm Maguire <address@hidden> writes:

> You need baby malloc due to the printf's called by debugging sbrk?
> Please let me know if for some reason you need baby_malloc still even
> when debugging versions are not used.

No, it's needed only for debugging output.

> Indeed!  Please keep us posted!  I'd wager that you're ready to go if
> you can unlimit the data seg size.

I found that the default settings (modifiable in /etc/login.conf) are
a soft limit of 64MB and a hard limit of 256MB.  Thus the following
patch, which increases the soft limit to the hard limit.

Attachment: txtjNhRlvGqoA.txt
Description: Text Data

The problem remains that if sbrk still fails (draconian sysadmin
lowered the limits?), there will be no helpful error messages, just
endless recursion and a core dump.  The following patch to main.c
replaces printf with write, thus avoiding the problem.

Attachment: txtyU3Vf35RLv.txt
Description: Text Data

It still doesn't work, though.  After a lot of output comes

>Can't write (saved_pre_gcl): errno 14

where 14 is EFAULT, "bad address".  I'll look at that later...

Regards,
Magnus

reply via email to

[Prev in Thread] Current Thread [Next in Thread]