PATCH (2.4.5): /dev/poll support

[Vincent Sweeney]

I think pgp-signing just barfed my last email (typical) so I'm retyping /
resending this:

Hi,
this patch adds Solaris 7/8 like /dev/poll support to the kernel.

I can claim no real credit for this as basically this is a fixed version of
a patch available from http://www.citi.umich.edu/projects/linux-scalability/
to compile correctly with 2.4.5 that only seemed to work with the 2.3.x
devel branch. The reason for this is so I can compile & test an application
on my home linux pc when I'm not around my nice work Solaris boxes :)

Please note, I have not got the knowledge of kernel development to know if
this patch is broken or badly written. It may be bugged and/or worse than
the standard poll() call but my application works so I'll leave profiling
etc to people more knowledgable than me.

Vince.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

PATCH (2.4.5): /dev/poll support

[Vincent Sweeney]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

more knowledgable than me can 

-BEGIN PGP SIGNATURE-
Version: PGPfreeware 7.0.3 for non-commercial use 

iQA/AwUBOzoyDz8f58rgPfa4EQIQRwCgxPE/JcNHuegj9fEklY7uhJr8O5UAnAtq
W46SBMHzCtsQYznIaKHFB8a+
=UaUj
-END PGP SIGNATURE-


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

PATCH (2.4.5): /dev/poll support

[Vincent Sweeney]

I think pgp-signing just barfed my last email (typical) so I'm retyping /
resending this:

Hi,
this patch adds Solaris 7/8 like /dev/poll support to the kernel.

I can claim no real credit for this as basically this is a fixed version of
a patch available from http://www.citi.umich.edu/projects/linux-scalability/
to compile correctly with 2.4.5 that only seemed to work with the 2.3.x
devel branch. The reason for this is so I can compile  test an application
on my home linux pc when I'm not around my nice work Solaris boxes :)

Please note, I have not got the knowledge of kernel development to know if
this patch is broken or badly written. It may be bugged and/or worse than
the standard poll() call but my application works so I'll leave profiling
etc to people more knowledgable than me.

Vince.


-
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: More 2.4 memory usage wierdness?

[Vincent Sweeney]

Vincent Sweeney wrote:
> 
> I have a question regarding a busy box I recently updated to
> '2.4.2-ac20'. It has been running for several hours without any real
> problems until I started getting 'dmesg' entries like:
> 
> Out of Memory: Killed process 30293 (httpd).
> Out of Memory: Killed process 32552 (mysqld).
> 
> Not a quick check with 'top' shows:
> 
> Mem:   771440K av,  396060K used,  375380K free,   0K shrd,4972K
> buff
> Swap:  379416K av,  345000K used,   34416K free  337128K
> cached
> 
> I have also attached my "/proc/meminfo" & "/proc/slabinfo" details.
> 
> Basically they all show the server has 370MB of free physical memory but
> its using lots of swap space and OOM is killing processes.
> 
> What gives?
> 
> Vince.

Maybe I should of also made it clear that is I run the same box with
2.2.18 I have no memory issues and swap is not used at all but when I
run with 2.4.2 I suddenly see my memory usage increase over 300MB and my
swap space maxes out.

-- 
Vincent Sweeney
System Architect
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

More 2.4 memory usage wierdness?

[Vincent Sweeney]

I have a question regarding a busy box I recently updated to
'2.4.2-ac20'. It has been running for several hours without any real
problems until I started getting 'dmesg' entries like:

Out of Memory: Killed process 30293 (httpd).
Out of Memory: Killed process 32552 (mysqld).

Not a quick check with 'top' shows:

Mem:   771440K av,  396060K used,  375380K free,   0K shrd,4972K
buff
Swap:  379416K av,  345000K used,   34416K free  337128K
cached

I have also attached my "/proc/meminfo" & "/proc/slabinfo" details.

Basically they all show the server has 370MB of free physical memory but
its using lots of swap space and OOM is killing processes.

What gives?

Vince.

-- 
Vincent Sweeney
System Architect
 meminfo
 slabinfo

More 2.4 memory usage wierdness?

[Vincent Sweeney]

I have a question regarding a busy box I recently updated to
'2.4.2-ac20'. It has been running for several hours without any real
problems until I started getting 'dmesg' entries like:

Out of Memory: Killed process 30293 (httpd).
Out of Memory: Killed process 32552 (mysqld).

Not a quick check with 'top' shows:

Mem:   771440K av,  396060K used,  375380K free,   0K shrd,4972K
buff
Swap:  379416K av,  345000K used,   34416K free  337128K
cached

I have also attached my "/proc/meminfo"  "/proc/slabinfo" details.

Basically they all show the server has 370MB of free physical memory but
its using lots of swap space and OOM is killing processes.

What gives?

Vince.

-- 
Vincent Sweeney
System Architect
 meminfo
 slabinfo

Re: More 2.4 memory usage wierdness?

[Vincent Sweeney]

Vincent Sweeney wrote:
 
 I have a question regarding a busy box I recently updated to
 '2.4.2-ac20'. It has been running for several hours without any real
 problems until I started getting 'dmesg' entries like:
 
 Out of Memory: Killed process 30293 (httpd).
 Out of Memory: Killed process 32552 (mysqld).
 
 Not a quick check with 'top' shows:
 
 Mem:   771440K av,  396060K used,  375380K free,   0K shrd,4972K
 buff
 Swap:  379416K av,  345000K used,   34416K free  337128K
 cached
 
 I have also attached my "/proc/meminfo"  "/proc/slabinfo" details.
 
 Basically they all show the server has 370MB of free physical memory but
 its using lots of swap space and OOM is killing processes.
 
 What gives?
 
 Vince.

Maybe I should of also made it clear that is I run the same box with
2.2.18 I have no memory issues and swap is not used at all but when I
run with 2.4.2 I suddenly see my memory usage increase over 300MB and my
swap space maxes out.

-- 
Vincent Sweeney
System Architect
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: cpqarray & 2.4.1+ hang

[Vincent Sweeney]

Marcus Meissner wrote:
> 
> Workaround: run the kernel with the 'noapic' option on its commandline.
> 
> The ServerWorks chipset used in this Compaq Server somehow does not pass
> the the relevant information to Linux mapping routines. :/
> 
> I have attached lspci -xxx and dmesg output of our DL360 below.
> 
> Ciao, Marcus

Yes that workaround means no hang on bootup anymore, thanks a lot.

---
Vincent Sweeney
[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

cpqarray & 2.4.1+ hang

[Vincent Sweeney]

I have a problem with the 2.4 series kernel running on a number of
Compaq ProLiant DL360 servers. The 2.2.x kernels and 2.4.0 work fine,
however from 2.4.1 onwards the boxes just hang at the following position
on bootup:

Partition check: 
  ida/c0d0:

I have also tested with 2.4.2-ac20 and the same problem occurs. Doing a
search on the web some people recommend changing the OS setting in the
Compaq BIOS to Linux fixes this problem, however my servers are already
running with this BIOS setting and I've also tested with numerous other
OS's in the BIOS but the same problem occurs.

Has anyone else fixed this problem with changing the BIOS OS? I've also
attached my .config file for extra details.

---
Vincent Sweeney
[EMAIL PROTECTED]
 config

Re: cpqarray 2.4.1+ hang

[Vincent Sweeney]

Marcus Meissner wrote:
 
 Workaround: run the kernel with the 'noapic' option on its commandline.
 
 The ServerWorks chipset used in this Compaq Server somehow does not pass
 the the relevant information to Linux mapping routines. :/
 
 I have attached lspci -xxx and dmesg output of our DL360 below.
 
 Ciao, Marcus

Yes that workaround means no hang on bootup anymore, thanks a lot.

---
Vincent Sweeney
[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/