Linux-Development-Sys Digest #350, Volume #8 Fri, 15 Dec 00 01:13:10 EST
Contents:
Re: Allocating memory for multiple pages ("Kang Mo Yang")
Intel StrongArm SA1110 processor ("Nils T. carlson")
SMP support and current->mm behaviour on 2.2 (Lee Cremeans)
Selective disabling of page cache for mmap ([EMAIL PROTECTED])
Re: Compiling C++ programs with GCC --> no GPL license implications (Austin Ziegler)
Re: Compiling C++ programs with GCC --> no GPL license implications (Austin Ziegler)
Re: Linux GUI programming (Eric Taylor)
Re: Compiling C++ programs with GCC --> no GPL license implications (Kaz Kylheku)
Re: Compiling C++ programs with GCC --> no GPL license implications (Peter Seebach)
Re: Compiling C++ programs with GCC --> no GPL license implications ("E. Robert
Tisdale")
Re: Compiling C++ programs with GCC --> no GPL license implications (Austin Ziegler)
Re: Compiling C++ programs with GCC --> no GPL license implications ("E. Robert
Tisdale")
Help for linux ext2 filesystem (Changxun Wu)
Where can I get documentation on BIOS int13 parameters? -nt ("Joshua Schaeffer")
is there a 2GB size limit on a file on NFS fielsystem on linux? (pete@x)
Re: is there a 2GB size limit on a file on NFS fielsystem on linux?
([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: "Kang Mo Yang" <[EMAIL PROTECTED]>
Subject: Re: Allocating memory for multiple pages
Date: Thu, 14 Dec 2000 15:39:40 -0500
Thanks, kmalloc() and get_free_pages() will do the job, but, I noticed that
there is a limit of 32 pages.
I need to allocte a size like 4MBytes.
Calling kmalloc() until it satisfy my memory requirement doesn't seems to
have solid promise neither.
Kang Mo,
Lee Ho <[EMAIL PROTECTED]> wrote in message
news:R5AZ5.1852$[EMAIL PROTECTED]...
> Kang Mo Yang Writes:
> >Is there any way of allocating a memory which is physically contiguous
for
> >multiple pages?
> >
> >I know that one way is disabling a part of memory and using the disabled
> >part from a driver. But, I am looking for a way more dynamic.
>
>
> In kernel module, kmalloc() and __get_free_pages() returns physically
continuous memory.
> If you want it in user application, allocate memory in kernel module,
> and remap it into virtual address space.
>
> *=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
> Lee, Ho. Software Engineer, Embedded Linux Dep, LinuxOne
> Mail : [EMAIL PROTECTED] (work), [EMAIL PROTECTED] (personal)
> Homepage : http://flyduck.com, http://linuxkernel.to
>
>
>
>
------------------------------
From: "Nils T. carlson" <[EMAIL PROTECTED]>
Subject: Intel StrongArm SA1110 processor
Date: Thu, 14 Dec 2000 15:41:29 -0600
Can anyone tell me if there is a embedded Linux implementation for the
StrongArm SA1110 processor.
I also need the ip networking package.
Of course, application development will be done using Linux on a
workstation and a cross-compiler.
If you know anything about this, please respond to this message or send
e-mail directly to:
[EMAIL PROTECTED]
Thanks,
Nils Carlson
Dallas, TX
------------------------------
From: Lee Cremeans <[EMAIL PROTECTED]>
Subject: SMP support and current->mm behaviour on 2.2
Date: Thu, 14 Dec 2000 23:47:35 GMT
Reply-To: [EMAIL PROTECTED]
I'm working on a driver that uses current->mm in a function that
ferrets out a physical address for a user-mode buffer. The function
seems to work fine on a 2.2 kernel without SMP support compiled in, but
it crashes the driver with SMP support there. I investigated and found
that current->mm was returning NULL. Is there something I'm missing?
This is my first time working with Linux's SMP support.
-lee
Sent via Deja.com
http://www.deja.com/
------------------------------
From: [EMAIL PROTECTED]
Subject: Selective disabling of page cache for mmap
Date: Fri, 15 Dec 2000 01:20:11 GMT
Hi,
I recently ran into a problem with the paging behavior for a very I/O
intensive server app.
My server app:
===========
My server performs a lot of reads (sustained 20-25MB/s) of read-only
requests from disk. In addition, the server uses a large in-memory
application cache (should be as large as 150-200 MB on a 256 MB
machine). The reads from disk are known to have little or no locality.
Blocks are "read" into memory by mmaping and mlocking them, to avoid
extra memory copies.
The behavior:
==========
The server starts paging significantly as the application cache fills
up. I observed anything from 200kB/s to 4 MB/s of paging. Paging
continues even if mlock the whole application and its cache. Calling
mlockall(MCL_CURRENT) every so often does not seem to help.
Calling mlockall(MCL_FUTURE) seems to slow down the paging a bit, but
not much. Generally, I see about 80% swapins and 20% swapouts.
I have checked that only the absolutely needed daemons are running
(sshd, kswapd, ... ). When paging, the cache size as reported by
top/vmstat is generous 80-90MB, enough to contain all binaries, the disk
blocks to be read, etc.
It seems to me that the problem may be that we cycle through the memory
very fast (we fill up a 100MB page cache in about 4-5 seconds), so
anything unlocked that had not been accessed for twice that amount of
time (if I understand the "clock" algorithm right) is subject to eviction.
Possible solution:
=============
a)
using swapoff -a, thus disabling all paging. That of course prevent
paging out to swap, but I was still observing a huge performance loss as
the application cache filled up. vmstat did not show any page acticvity.
I assume that part of my text segment could just be dropped and then
just be loaded from the underlying file again. As vmstat shows no swap
in's, I'm wondering if this situation just does not occur, or if those
page faults just are not reported as "swap ins"
b)
would it be possible to introduce a new flag to mmap(), that will
prevent the pages being mapped from being put in the page cache
(something alike MAP_ANONYMOUS, just that it reads the blocks from disk
on a page fault) ?
I would appreciate if anyone can offer any ideas on why the server
starts paging at all, and if you see any other solutions than a) or b),
or if you have know an implementation to b) and can point me to it.
Thanks a bunch!!
Max
Sent via Deja.com
http://www.deja.com/
------------------------------
Crossposted-To: comp.lang.c++,gnu.misc.discuss
From: Austin Ziegler <[EMAIL PROTECTED]>
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
Date: Thu, 14 Dec 2000 21:47:33 -0500
On Thu, 14 Dec 2000, Mike Stump wrote:
> In article <[EMAIL PROTECTED]>, jbs <[EMAIL PROTECTED]> wrote:
>> That's probably not the biggest difference. The biggest difference
>> is that open source licenses are enforced largely by community
>> consensus and peer pressure.
> I agree. And in fact, I like a world where people can operate this
> way, and don't need to go to court to work together. Some companies
> might not mind loosing a $10,000 judgement in a copyright case, but
> might not like a $100,000 loss in good will that not conforming to
> peer pressure and working with the community might bring. Then again,
> I may be an eternal optimist.
>
> Apple lost good will with the FSF and look where it got them.
> :-)
Doesn't seem to have hurt them at all. There are other things that hurt
them, but pissing off the FSF is one of my smallest worries, and is
likely to be Apple's, too.
I also find your 'dream world' a nightmare -- peer pressure shouldn't
have anything to do with why one shares one's code, and admitting that
it's required indicates a lot about the so-called 'free' software
movement.
Give your code away because you want to, not because some pissant
writes a licence that tells you that you have to.
-f
--
austin ziegler * fant0me(at)the(dash)wire(d0t)c0m * Ni bhionn an rath ach
ICQ#25o49818 (H) * aziegler(at)s0lect(d0t)c0m * mar a mbionn an smacht
ICQ#21o88733 (W) * fant0me526(at)yah00(d0t)c0m * (There is no Luck
AIM Fant0me526 *-s/0/o/g--------&&--------s/o/0/g-* without Discipline)
Toronto.ON.ca * I speak for myself alone *-----------------------
------------------------------
Crossposted-To: comp.lang.c++,gnu.misc.discuss
From: Austin Ziegler <[EMAIL PROTECTED]>
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
Date: Thu, 14 Dec 2000 21:49:05 -0500
On Thu, 14 Dec 2000, jbs wrote:
> Mike Stump wrote:
>> Apple lost good will with the FSF and look where it got them.
>> :-)
> It probably got them knocked down harder than you might think. I
> believe that the cold shoulder Apple has gotten from the free software
> community with their Darwin project is partially due to remnants bad
> will from the days of the FSF boycott, along with the defects in their
> license which the FSF and free software community has identified.
>
> Play nice and you get cooperation. Step out of line and you get
> ostracized.
I'm so glad that you're there to tell us what playing nice is ... good
thing that there's the rule of law to prevent tyrants like you from running
things into the ground (and the same for corporations).
-f
--
austin ziegler * fant0me(at)the(dash)wire(d0t)c0m * Ni bhionn an rath ach
ICQ#25o49818 (H) * aziegler(at)s0lect(d0t)c0m * mar a mbionn an smacht
ICQ#21o88733 (W) * fant0me526(at)yah00(d0t)c0m * (There is no Luck
AIM Fant0me526 *-s/0/o/g--------&&--------s/o/0/g-* without Discipline)
Toronto.ON.ca * I speak for myself alone *-----------------------
------------------------------
From: Eric Taylor <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.development
Subject: Re: Linux GUI programming
Date: Fri, 15 Dec 2000 02:54:27 GMT
I was faced with learning a GUI again (after learning
a few including the AMIGA) I wanted this to be my last
GUI.
I chose C/tcl/tk combinations as these can be very
portable and tcl/tk has a very nice GUI api. If I can
write the entire program in tcl/tk, I do so. If tcl is
not sufficient, then I add C code. Either way, same GUI api.
And since tcl is a scripting language, I can add scripting
to a GUI application with ease.
I do use some tools such as mktclapp to glue tcl/tk to C code. And,
I use visual tcl to build my GUIs. All these tools are open source.
If I am going to include C code, I then usually use ms vis c/c++ to
develop
and port to linux regularly so nothing non-portable creeps in.
The vis c help command provides excellent documentation on
what is ansi standard so I don't use any microsoft specific code.
I've built 5-6 programs this way so far, and I ain't ever
going back to x-windows or m.s. foundation classes.
Eric
Joseph Virzi wrote:
>
> I would like to write a GUI based application, just like I can with
> Windows. As an application and device driver writer, I have been forced
> to use Windows because of the GUI. In all other respects, Linux is
> superior.
>
> I have done some research, and all I can find reference to are functions
> to plot polygons, fills, etc.
>
> I would like functions such as CreateWindow(), SetFocus() and
> GetEvents() so that I can control my application with my mouse as it
> becomes to wieldy with my keyboard.
>
> Surely this exists. Does anyone know of a package that might look and
> feel just like Windows GUI programming. On GUI, from what I have seen so
> far, I thnk Windows is superior. It doesn't have to be that way.
>
> -Joe
------------------------------
From: [EMAIL PROTECTED] (Kaz Kylheku)
Crossposted-To: comp.lang.c++,gnu.misc.discuss
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
Reply-To: [EMAIL PROTECTED]
Date: Fri, 15 Dec 2000 02:54:58 GMT
On Thu, 14 Dec 2000 21:47:33 -0500, Austin Ziegler <[EMAIL PROTECTED]>
wrote:
>On Thu, 14 Dec 2000, Mike Stump wrote:
>> Apple lost good will with the FSF and look where it got them.
>> :-)
>
>Doesn't seem to have hurt them at all. There are other things that hurt
>them, but pissing off the FSF is one of my smallest worries, and is
>likely to be Apple's, too.
>
>I also find your 'dream world' a nightmare -- peer pressure shouldn't
>have anything to do with why one shares one's code, and admitting that
>it's required indicates a lot about the so-called 'free' software
>movement.
>
>Give your code away because you want to, not because some pissant
>writes a licence that tells you that you have to.
You mean like some pissant from some company that wants to jump on the
open sourc bandwagon to reap the development benefits, while keeping
the software as proprietary as possible? Damn the FSF for opposing
that!
------------------------------
Crossposted-To: comp.lang.c++,gnu.misc.discuss
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
From: [EMAIL PROTECTED] (Peter Seebach)
Date: 15 Dec 2000 03:39:41 GMT
In article <[EMAIL PROTECTED]>, Mike Stump <[EMAIL PROTECTED]> wrote:
>If what you say is true, please provide evidence of a single case that
>has been litigated and won, where someone else besides the FSF sued.
>I bet you cannot.
Has the FSF ever actually won a case, as opposed to getting a settlement,
on this issue?
-s
--
Copyright 2000, All rights reserved. Peter Seebach / [EMAIL PROTECTED]
C/Unix wizard, Pro-commerce radical, Spam fighter. Boycott Spamazon!
Consulting & Computers: http://www.plethora.net/
------------------------------
From: "E. Robert Tisdale" <[EMAIL PROTECTED]>
Crossposted-To: comp.lang.c++,gnu.misc.discuss
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
Date: Fri, 15 Dec 2000 03:54:20 +0000
jbs wrote:
> That's probably not the biggest difference.
> The biggest difference is that
> open source licenses are enforced
> largely by community consensus and peer pressure.
>
> Of course, this would mean that informed opinions on a newsgroup
> might well be *more valuable* than the "sound legal advice" god
> that Becker worships, or the "anybody can sue you" bogeyman
> that Tisdale dreads.
You know, until just a couple of days ago,
I thought that voters -- and not the courts -- would decide
who would be President of the United States.
------------------------------
Crossposted-To: comp.lang.c++,gnu.misc.discuss
From: Austin Ziegler <[EMAIL PROTECTED]>
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
Date: Thu, 14 Dec 2000 23:08:39 -0500
On Fri, 15 Dec 2000, Kaz Kylheku wrote:
> On Thu, 14 Dec 2000 21:47:33 -0500, Austin Ziegler <[EMAIL PROTECTED]>
> wrote:
>> On Thu, 14 Dec 2000, Mike Stump wrote:
>>> Apple lost good will with the FSF and look where it got them.
>>> :-)
>> Doesn't seem to have hurt them at all. There are other things that hurt
>> them, but pissing off the FSF is one of my smallest worries, and is
>> likely to be Apple's, too.
>> I also find your 'dream world' a nightmare -- peer pressure shouldn't
>> have anything to do with why one shares one's code, and admitting that
>> it's required indicates a lot about the so-called 'free' software
>> movement.
>> Give your code away because you want to, not because some pissant
>> writes a licence that tells you that you have to.
> You mean like some pissant from some company [...]
Nope. Please work on your reading comprehension skills. I said no such
thing.
-f
--
austin ziegler * fant0me(at)the(dash)wire(d0t)c0m * Ni bhionn an rath ach
ICQ#25o49818 (H) * aziegler(at)s0lect(d0t)c0m * mar a mbionn an smacht
ICQ#21o88733 (W) * fant0me526(at)yah00(d0t)c0m * (There is no Luck
AIM Fant0me526 *-s/0/o/g--------&&--------s/o/0/g-* without Discipline)
Toronto.ON.ca * I speak for myself alone *-----------------------
------------------------------
From: "E. Robert Tisdale" <[EMAIL PROTECTED]>
Crossposted-To: comp.lang.c++,gnu.misc.discuss
Subject: Re: Compiling C++ programs with GCC --> no GPL license implications
Date: Fri, 15 Dec 2000 04:05:49 +0000
Mike Stump wrote:
> In article <[EMAIL PROTECTED]>,
> E. Robert Tisdale <[EMAIL PROTECTED]> wrote:
> >The FSF probably is NOT going to sue anybody even if they modify
> >GPL'd software and refuse to distribute the modified source code
> >along with the binaries.
>
> This and a dollar will get you sued, be my guest, try it, let us all
> know the outcome.
>
> >The FSF just can't afford to do that.
>
> Sure they can.
>
> >If you want the modified source code you will probably need to sue
> >them yourself.
>
> Ok, you want to go that route, quote the law the permits you to win
> such a suit. If you cannot quote it, describe it in detail as best
> you can, and we can see if others can help identify it.
Once again, I'm not a lawyer.
I can't advise you or anyone else on the law.
I am not a judge. I can't decide your suit.
If you want to continue this discussion,
please move it to one of the legal newsgroups
where it might be appreciated.
It is off-topic in these newsgroups.
------------------------------
From: Changxun Wu <[EMAIL PROTECTED]>
Subject: Help for linux ext2 filesystem
Date: Thu, 14 Dec 2000 23:37:56 -0500
In kernel 2.2.17, fs/ext2/namei.c, the ext2_add_entry() function, it
seems that whenever we find a big enough hole in the directory's data
block, we insert the new entry into it. However, how can we be
sure that we don't insert a duplicate file with the same name? It seems
that we just check if it has
some already existed file when we advance our pointer to current dentry.
So, if I insert the new entry
at this point, how can I be sure that there won't be another file
sharing the same name in the data block
area we haven't scanned yet? Thanks a lot!
if (ext2_match (namelen, name, de)) {
*err = -EEXIST;
brelse (bh);
return NULL;
}
if ((le32_to_cpu(de->inode) == 0 && le16_to_cpu(de->rec_len) >=
rec_len) ||
(le16_to_cpu(de->rec_len) >= EXT2_DIR_REC_LEN(de->name_len) +
rec_len)) {
offset += le16_to_cpu(de->rec_len);
if (le32_to_cpu(de->inode)) {
de1 = (struct ext2_dir_entry_2 *) ((char *) de +
EXT2_DIR_REC_LEN(de->name_len));
de1->rec_len = cpu_to_le16(le16_to_cpu(de->rec_len) -
EXT2_DIR_REC_LEN(de->name_len));
de->rec_len = cpu_to_le16(EXT2_DIR_REC_LEN(de->name_len));
de = de1;
}
de->inode = 0;
de->name_len = namelen;
de->file_type = 0;
memcpy (de->name, name, namelen);
/*
* XXX shouldn't update any times until successful
* completion of syscall, but too many callers depend
* on this.
*
* XXX similarly, too many callers depend on
* ext2_new_inode() setting the times, but error
* recovery deletes the inode, so the worst that can
* happen is that the times are slightly out of date
* and/or different from the directory change time.
*/
dir->i_mtime = dir->i_ctime = CURRENT_TIME;
dir->u.ext2_i.i_flags &= ~EXT2_BTREE_FL;
mark_inode_dirty(dir);
dir->i_version = ++event;
mark_buffer_dirty(bh, 1);
*res_dir = de;
*err = 0;
return bh;
}
------------------------------
From: "Joshua Schaeffer" <[EMAIL PROTECTED]>
Subject: Where can I get documentation on BIOS int13 parameters? -nt
Date: Fri, 15 Dec 2000 04:41:12 GMT
------------------------------
From: pete@x <[EMAIL PROTECTED]>
Subject: is there a 2GB size limit on a file on NFS fielsystem on linux?
Date: 14 Dec 2000 20:29:50 -0800
hello,
I have a linux box connected to a solaris box.
I mount an NFS mount point on Solaris from Linux,
so from Linux, I am able to access Solaris file over nfs.
Can I create a file > 2GB on this new mounted file system now?
I know ext2 has a limit of 2GB (this is on intel hw).
Also I know that solaris itself can support > 2GB size.
But since I'll be using tar command from linux, and tar itself
could not support > 2 GB, even if the underlying filesystem support
>2GB (nfs in this case), it might still not work ?
thanks,
pete
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: is there a 2GB size limit on a file on NFS fielsystem on linux?
Date: Fri, 15 Dec 2000 05:52:11 GMT
pete@x <[EMAIL PROTECTED]> writes:
> I have a linux box connected to a solaris box.
> I mount an NFS mount point on Solaris from Linux,
> so from Linux, I am able to access Solaris file over nfs.
>
> Can I create a file > 2GB on this new mounted file system now?
> I know ext2 has a limit of 2GB (this is on intel hw).
>
> Also I know that solaris itself can support > 2GB size.
>
> But since I'll be using tar command from linux, and tar itself
> could not support > 2 GB, even if the underlying filesystem support
> >2GB (nfs in this case), it might still not work ?
NFS version 2, which is all that is fully supported, as of yet, only
supports 2GB files as its maximum.
Which means that the fact that ext2 supports 2TB files, as do modern
versions of Linux on IA-32, as does TAR, if compiled with the LFS
option, is _irrelevant_ because the limitation is with NFS.
--
(reverse (concatenate 'string "ac.notelrac.teneerf@" "454aa"))
<http://www.ntlug.org/~cbbrowne/>
"The problem with X is that it's overadequate" -- Dennis Ritchie
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list by posting to the
comp.os.linux.development.system newsgroup.
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Development-System Digest
******************************
