zsh-2.6beta20 doesn't compile under Linux pre2.0.14

zsh-2.6beta20 doesn't compile under Linux pre2.0.14

[jkroeger]

Recent Linux kernels introduced a new limit for the address space of a process:

#define RLIMIT_AS       9               /* address space limit */

This makes compiling zsh fail with the following error message:

builtin.c:3539: #error there are some unknown limits.  Fix me!
make[1]: *** [builtin.o] Error 1
make[1]: Leaving directory `/usr/src/bin/zsh-2.6-beta20/Src'
make: *** [Src] Error 2

Unfortunately I don't know how to fix zsh so that it can handle the new
resource limit :-(

-- 
Johannes Kroeger		<jkroeger@squirrel.owl.de>
Send me mail with subject "send pgp-key" to get my PGP key

Re: zsh-2.6beta20 doesn't compile under Linux pre2.0.14

[Zoltan Hidvegi]

> Recent Linux kernels introduced a new limit for the address space of a process:
> 
> #define RLIMIT_AS       9               /* address space limit */
> 
> This makes compiling zsh fail with the following error message:
> 
> builtin.c:3539: #error there are some unknown limits.  Fix me!
> make[1]: *** [builtin.o] Error 1
> make[1]: Leaving directory `/usr/src/bin/zsh-2.6-beta20/Src'
> make: *** [Src] Error 2
> 
> Unfortunately I don't know how to fix zsh so that it can handle the new
> resource limit :-(

It is surprising that the there days before the final 2.0 release a new
feature appeared.  Here is the patch.  I think zsh will be the first shell
which can set this new limit.  There is no patch to the manual simply
because I do not know what is this limit.  Also a new flag should be
inveted to set this limit with ulimit.

Please tell me your suggestions.

Zoltan

 *** Src/builtin.c	1996/06/06 17:23:43	2.39
 --- Src/builtin.c	1996/06/10 16:53:08
 ***************
 *** 3535,3540 ****
 --- 3535,3545 ----
   #  define NEXT_RLIM (RLIMIT_MEMLOCK + 1)
       "memorylocked",
   # endif /* RLIMIT_MEMLOCK */
 + # if defined RLIMIT_AS && RLIMIT_AS == NEXT_RLIM
 + #  undef NEXT_RLIM
 + #  define NEXT_RLIM (RLIMIT_AS + 1)
 +     "addressspace",
 + # endif /* RLIMIT_AS */
   # if NEXT_RLIM != RLIM_NLIMITS
      #error there are some unknown limits.  Fix me!
   # endif
 ***************
 *** 3957,3962 ****
 --- 3962,3973 ----
   	limit /= 1024;
   	break;
   # endif /* RLIMIT_VMEM */
 + # ifdef RLIMIT_AS
 +     case RLIMIT_AS:
 + 	printf("address space (kb)         ");
 + 	limit /= 1024;
 + 	break;
 + # endif /* RLIMIT_AS */
       }
       /* display the limit */
       if (limit == RLIM_INFINITY)

Re: zsh-2.6beta20 doesn't compile under Linux pre2.0.14

[Steven L Baur]

>>>>> "Zoltan" == Zoltan Hidvegi <hzoli@cs.elte.hu> writes:

(Your patch produces a compilable zsh on Linux 2.0, thanks).

Zoltan> It is surprising that the there days before the final 2.0
Zoltan> release a new feature appeared.  Here is the patch.  I think
Zoltan> zsh will be the first shell which can set this new limit.
Zoltan> There is no patch to the manual simply because I do not know
Zoltan> what is this limit.  Also a new flag should be inveted to set
Zoltan> this limit with ulimit.

It's related to the problem found recently where ``grep x /dev/zero''
would crash the system.  I don't believe it does anything, at least I
haven't found any code via grep in the kernel source that does
anything with it.

Zoltan> Please tell me your suggestions.

Here's the only authoritative information about it I'm aware of:
X-From-Line: owner-linux-kernel-outgoing@vger.rutgers.edu  Fri Jun  7 17:17:37 1996
Return-Path: owner-linux-kernel-outgoing@vger.rutgers.edu
Received: from wolverine.hq.cic.net (work.nether.net [198.87.18.3]) by deanna.miranova.com (8.7.5/8.6.9) with ESMTP id RAA23647 for <steve@miranova.com>; Fri, 7 Jun 1996 17:17:36 -0700
Received: from vger.rutgers.edu (vger.rutgers.edu [128.6.190.2]) by wolverine.hq.cic.net (8.7.5/8.6.9+jared) with ESMTP id UAA12029; Fri, 7 Jun 1996 20:08:13 -0400
Received: by vger.rutgers.edu id <106459-20973>; Fri, 7 Jun 1996 07:46:19 -0500
Date: 	Fri, 7 Jun 1996 09:22:26 +0300 (EET DST)
From: Linus Torvalds <torvalds@cs.helsinki.fi>
To: Michael Meskes <meskes@informatik.rwth-aachen.de>
cc: Linux-Kernel Mailinglist <linux-kernel@vger.rutgers.edu>
Subject: Re: the mmap() problem, a'la "grep x /dev/zero"
In-Reply-To: <199606070607.IAA00447@circe.informatik.rwth-aachen.de>
Message-ID: <Pine.LNX.3.91.960607091748.10892A-100000@linux.cs.Helsinki.FI>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-linux-kernel@vger.rutgers.edu
Precedence: bulk
Lines: 33
Xref: deanna.miranova.com linux-kernel:1385



On Fri, 7 Jun 1996, Michael Meskes wrote:
> 
> circe:meskes 105) grep x /dev/zero
> grep: memory exhausted

Umm, this looks ok - that's what supposed to happen. Linux can't fix the 
fact that grep doesn't work well on /dev/null (or any file that is that 
huge), but at least the above isn't a crash..

> It allocated much more memory than it should. I had a limit of 4MB, but the
> size went up to 12MB (5MB resident).

Umm, _what_ size? If you set a RLIMIT_DATA (or whatever the name is), 
that will only make a difference for the process bss (*), and if grep 
uses mmap(), then we'd need to have a "virtual memory" limit. However, we 
don't have any such limit, although I added a RLIMIT_AS for 2.0 (but it's 
currently unchecked).

(*) that's the curent Linux interpretation, at least. Do other unixes 
consider mmap'ed data part of RLIMIT_DATA?

>	 			Also I was able to run the memtest
> program on 8MB despite datasize limited to 1MB. Hmm I have to check the
> sources to see what happened. And my program with a global definition of
> 'char a[128MB]' still dumpos core.

Well, dumping core is the corect thing to do - it failed to get the BSS 
at exec time, so when the program actually ran it was "kind of" unhappy..

		Linus


-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be proofread for $250/hour.
Andrea Seastrand: For your vote on the Telecom bill, I will vote for anyone
except you in November.

Re: zsh-2.6beta20 doesn't compile under Linux pre2.0.14

[Johannes Kroeger]

Hello Zoltan,

Thank you for your quick patch.

As Steven L Baur <steve@miranova.com> noticed in
<m2hgsjimn9.fsf@deanna.miranova.com>, RLIMIT_AS is only defined
in <asm/resource.h> but not yet used in the kernel.

>[...].  Also a new flag should be invented to set this limit with ulimit.
>Please tell me your suggestions.

RLIMIT_AS is only defined for the Intel, Alpha and Mips architectures.

<asm-i386/resource.h> and <asm-alpha/resource.h> have no RLIMIT_VMEM,
and <asm-mips/resource.h> has the following definition:

#define RLIMIT_VMEM     6               /* mapped memory */
#define RLIMIT_AS       RLIMIT_VMEM

So it seems that 'ulimit -v' and 'limit vmemorysize' are the right commands
to control the new resource limit.
Anyway, I would recommend you to ask the gurus in the linux-kernel list
about the intended implementation of the "address space limit".


Cheers,
Johannes

-- 
Johannes Kroeger		<jkroeger@squirrel.owl.de>
Send me mail with subject "send pgp-key" to get my PGP key

Re: zsh-2.6beta20 doesn't compile under Linux pre2.0.14

[Felix von Leitner]

Thus spake jkroeger@squirrel.owl.de (jkroeger@squirrel.owl.de):

> Recent Linux kernels introduced a new limit for the address space of a process:

> #define RLIMIT_AS       9               /* address space limit */

> This makes compiling zsh fail with the following error message:

> builtin.c:3539: #error there are some unknown limits.  Fix me!
> make[1]: *** [builtin.o] Error 1
> make[1]: Leaving directory `/usr/src/bin/zsh-2.6-beta20/Src'
> make: *** [Src] Error 2

Apply this diff, it works for me.

Felix

--- Src/builtin.c	Thu Jun 13 02:23:09 1996
+++ Src/builtin.c.new	Thu Jun 13 02:22:49 1996
@@ -3535,6 +3535,11 @@
 #  define NEXT_RLIM (RLIMIT_MEMLOCK + 1)
     "memorylocked",
 # endif /* RLIMIT_MEMLOCK */
+# if defined RLIMIT_AS && RLIMIT_AS == NEXT_RLIM
+#  undef NEXT_RLIM
+#  define NEXT_RLIM (RLIMIT_AS + 1)
+    "addressspace",
+# endif /* RLIMIT_AS */
 # if NEXT_RLIM != RLIM_NLIMITS
    #error there are some unknown limits.  Fix me!
 # endif