linuxconsole-dev

ruby 2.6.10

[Aivils S. <ai...@la...>]

Hi All!

ruby is synced to 2.6.10 without frambuffer support.
http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050108.diff.bz2

Seems i drop framebuffer support some kernel versions.
I won't see serial console output anymore, but Antonio Daplas
rewrite 60% of code of fbcon each kernel version.

Aivils

Re: ruby 2.6.10

[Zoltan B. <zb...@fr...>]

Aivils Stoss =EDrta:
> Hi All!
>=20
> ruby is synced to 2.6.10 without frambuffer support.
> http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050108.diff.bz2
>=20
> Seems i drop framebuffer support some kernel versions.
> I won't see serial console output anymore, but Antonio Daplas
> rewrite 60% of code of fbcon each kernel version.

I somehow expected this. About the same amount of changes went
mainstream after 2.6.10 was released, too.

And when will it go into the CVS?

Best regards,
Zolt=E1n B=F6sz=F6rm=E9nyi

Re: ruby 2.6.10

[Aivils S. <ai...@la...>]

On Sat, 8 Jan 2005, Zoltan Boszormenyi wrote:

> Aivils Stoss =EDrta:
> > Hi All!
> >=20
> > ruby is synced to 2.6.10 without frambuffer support.
> > http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050108.diff.bz2
> >=20
> > Seems i drop framebuffer support some kernel versions.
> > I won't see serial console output anymore, but Antonio Daplas
> > rewrite 60% of code of fbcon each kernel version.
>=20
> I somehow expected this. About the same amount of changes went
> mainstream after 2.6.10 was released, too.
>=20
> And when will it go into the CVS?
>=20

I am away. I can't recall my sf.net account password :o(

Aivils

Re: ruby 2.6.10

[James v. Z. <ja...@dv...>]

2.6.10-ruby is still subject to the inlining compile failure under fc3 /
gcc3.4.2

Perhaps some of the kernel hackers and experienced coders here might be
able to explain what needs to be changed so this can compile under fc3
and others using gcc3.4.2 or therabouts?

I noted previously that the 269 ruby patch applies and compiles against
the FC3 kernel source tree. It does not compile under FC3 against the
stock 269 kernel tree. Although the ruby patched FC3 kernel compiles, it
has not behaved stable for me.

Is there a compiler flag that can be changed to prevent this compile
problem?

Or should we consider ruby incompatible with FC3 and/or this gcc version
in the meantime?

James

On Sun, 2005-01-09 at 03:19, Aivils Stoss wrote:
> Hi All!
> 
> ruby is synced to 2.6.10 without frambuffer support.
> http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050108.diff.bz2
> 
> Seems i drop framebuffer support some kernel versions.
> I won't see serial console output anymore, but Antonio Daplas
> rewrite 60% of code of fbcon each kernel version.
> 
> Aivils
> 
> 
> 
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by: Beat the post-holiday blues
> Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
> It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> Linuxconsole-dev mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxconsole-dev
> 

Re: ruby 2.6.10

[Zoltan B. <zb...@fr...>]

James van Zeeland =EDrta:
> 2.6.10-ruby is still subject to the inlining compile failure under fc3 =
/
> gcc3.4.2
>=20
> Perhaps some of the kernel hackers and experienced coders here might be
> able to explain what needs to be changed so this can compile under fc3
> and others using gcc3.4.2 or therabouts?
>=20
> I noted previously that the 269 ruby patch applies and compiles against
> the FC3 kernel source tree. It does not compile under FC3 against the
> stock 269 kernel tree. Although the ruby patched FC3 kernel compiles, i=
t
> has not behaved stable for me.
>=20
> Is there a compiler flag that can be changed to prevent this compile
> problem?
>=20
> Or should we consider ruby incompatible with FC3 and/or this gcc versio=
n
> in the meantime?
>=20
> James

Apply this patch after you applied ruby-2.6.10-nofb-20050108
on linux-2.6.10. I solved it the ame way when I made my 2.6.10-rc3-ruby
patch. (Unfortunately that never made it to the list, the attachments
made my mail too large. It's still in the moderator's queue.)
I am also on FC3. This patch also contains two warning fixes, one in
fs/compat_ioctl.c and one in drivers/char/vt.c.

Best regards,
Zolt=E1n B=F6sz=F6rm=E9nyi

Re: ruby 2.6.10

[Aivils S. <ai...@la...>]

On Sun, 9 Jan 2005, Zoltan Boszormenyi wrote:

> James van Zeeland =EDrta:
> > 2.6.10-ruby is still subject to the inlining compile failure under fc3 =
/
> > gcc3.4.2
> >=20
>=20
> Apply this patch after you applied ruby-2.6.10-nofb-20050108
> on linux-2.6.10. I solved it the ame way when I made my 2.6.10-rc3-ruby
> patch. (Unfortunately that never made it to the list, the attachments
> made my mail too large. It's still in the moderator's queue.)
> I am also on FC3. This patch also contains two warning fixes, one in
> fs/compat_ioctl.c and one in drivers/char/vt.c.
>=20
> Best regards,
> Zolt=E1n B=F6sz=F6rm=E9nyi
>=20

Zoltan's patches applayed
http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050110.diff.bz2

Aivils

Candidate multi-console friendly 2.6 scheduling?

[James v. Z. <ja...@dv...>]

Zoltan's patches applayed
> http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050110.diff.bz2
> 
> Aivils

Now have a working FC3.
Thanks for your work on ruby!

http://kerneltrap.org/node/4493
I saw this and couldn't help but wonder how it might perform,
particularly with the "unusual" load of multiple local consoles.
Towards that end I have applied the zaphod scheduler and gene algorythm
to my 2.6.9 ruby kernel. I have noticed a real change in CPU usage
behaviour. I've observed an adaptec hardware RAID array increase
throughput and other little improvements in behaviour. Reduced idle
load...

http://www.members.westnet.com.au/vanzeeland/Linux/kernel/patch-2.6.9-ruby.vz9x1.diff.gz
Here 'tis. Includes a lot of driver updates and other functionals too.
patch against kernel.org 2.6.9

I am wondering if this scheduler configuration might improve response
for multi console gaming and multi-console loads in general? The task
and anticipatory I/O scheduler tuneable variables will converge towards
the optimum settings for whatever load happens to be running. When the
load changes the schedulers converge towards a new optimum.

I am very curious to hear others thoughts. I've had a couple of days
uptime with it now; doesn't seem to have effected stability at all, and
seeing modest performance improvements. Notably quake3 was very playable
while recompiling a kernel; even on this dual Athlon playability would
previously suffer much more doing this.

Trying two heads of Unreal Tournament 2004 demo, one head in spectator
more chasing the other head around; bear in mind this is a dual 2800+
athlon with 2GbDDR, it took much longer to load the second instance, I
got good agp performance at 1024x768 and PCI at 640x480. The PCI
performance seemed OK, but nothing to write home about, bit choppy.

Quake3 showed improvements with two consoles at 1024x768. Good
performance on both. Previously both had worked ok but there was
noticable moments of lag but playable. Now both are smooth with no laggy
moments and I think I have a higher framerate, at least on the AGP
console.

James

P.S Trying to take it further, I would like to try this gene algorythm
kernel on some single CPU FC3 workstations; I tried to patch the ruby
inlining fixes to 2.6.9, however the following hunk failed, and I know
not how to fix it:
***************
*** 1372,1378 ****
        console_conditional_schedule();
        release_console_sem();

- out:
        return n;
  #undef FLUSH
  }
--- 1372,1377 ----
        console_conditional_schedule();
        release_console_sem();

        return n;
  #undef FLUSH
  }

is there a simple fix?

Re: Candidate multi-console friendly 2.6 scheduling?

[Hugo V. <hvw...@ya...>]

--- James van Zeeland <ja...@dv...> wrote:

> 
> Zoltan's patches applayed
> >
>
http://www.ltn.lv/~aivils/files/ruby-2.6.10-nofb-20050110.diff.bz2
> > 
> > Aivils
> 
> Now have a working FC3.
> Thanks for your work on ruby!
> 
> http://kerneltrap.org/node/4493
> I saw this and couldn't help but wonder how it might
> perform,
> particularly with the "unusual" load of multiple
> local consoles.
> Towards that end I have applied the zaphod scheduler
> and gene algorythm
> to my 2.6.9 ruby kernel. I have noticed a real
> change in CPU usage
> behaviour. I've observed an adaptec hardware RAID
> array increase
> throughput and other little improvements in
> behaviour. Reduced idle
> load...
> 
>
http://www.members.westnet.com.au/vanzeeland/Linux/kernel/patch-2.6.9-ruby.vz9x1.diff.gz
> Here 'tis. Includes a lot of driver updates and
> other functionals too.
> patch against kernel.org 2.6.9

Does that already include the vesafb-tng patch?

hugo








		
__________________________________ 
Do you Yahoo!? 
The all-new My Yahoo! - Get yours free! 
http://my.yahoo.com 
 

Re: Candidate multi-console friendly 2.6 scheduling?

[Hugo V. <hvw...@ya...>]

--- Hugo Vanwoerkom <hvw...@ya...> wrote:

> 
> Does that already include the vesafb-tng patch?
> 

Thanks, it does.
But even with that and similar kernel options, I get a
difference of 50% cpu usage between 2.6.9 and 2.4.29
in playing xine-ui.

Also if you change line 44392 of the referenced diff
file then you have pc speaker pich back and EVERYBODY
was waiting for that wonderful feature with avid
breath ;-)

Hugo


		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - 250MB free storage. Do more. Manage less. 
http://info.mail.yahoo.com/mail_250

Re: Candidate multi-console friendly 2.6 scheduling?

[Hugo V. <hvw...@ya...>]

--- Hugo Vanwoerkom <hvw...@ya...> wrote:

> 
> --- Hugo Vanwoerkom <hvw...@ya...> wrote:
> 
> > 
> > Does that already include the vesafb-tng patch?
> > 
> 
> Thanks, it does.
> But even with that and similar kernel options, I get
> a
> difference of 50% cpu usage between 2.6.9 and 2.4.29
> in playing xine-ui.
> 
> Also if you change line 44392 of the referenced diff
> file 

sorry: by changing the "if" into an "else if"


then you have pc speaker pitch back and
> EVERYBODY
> was waiting for that wonderful feature with avid
> breath ;-)
> 
> Hugo
> 
> 
> 		
> __________________________________ 
> Do you Yahoo!? 
> Yahoo! Mail - 250MB free storage. Do more. Manage
> less. 
> http://info.mail.yahoo.com/mail_250
> 
> 
>
-------------------------------------------------------
> The SF.Net email is sponsored by: Beat the
> post-holiday blues
> Get a FREE limited edition SourceForge.net t-shirt
> from ThinkGeek.
> It's fun and FREE -- well,
> almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> Linuxconsole-dev mailing list
> Lin...@li...
>
https://lists.sourceforge.net/lists/listinfo/linuxconsole-dev
> 



		
__________________________________ 
Do you Yahoo!? 
The all-new My Yahoo! - What will yours do?
http://my.yahoo.com 

Re: Candidate multi-console friendly 2.6 scheduling?

[Aivils S. <ai...@la...>]

On Tue, 11 Jan 2005, Hugo Vanwoerkom wrote:

> 
> --- Hugo Vanwoerkom <hvw...@ya...> wrote:
> 
> > 
> > Does that already include the vesafb-tng patch?
> > 
> 
> Thanks, it does.
> But even with that and similar kernel options, I get a
> difference of 50% cpu usage between 2.6.9 and 2.4.29
> in playing xine-ui.

top LIES! top must be "tuned" according system timer.
under 2.4.XX system timer 100Hz. You can force
top show satisfactory values if You set 2.6.XX system
timer to 100, default is 1000
include/asm/param.h:5:# define HZ               1000

check out procps package updates :o)

Aivils

Re: Candidate multi-console friendly 2.6 scheduling?

[Hugo V. <hvw...@ya...>]

--- Aivils Stoss <ai...@la...> wrote:

> top LIES! top must be "tuned" according system
> timer.
> under 2.4.XX system timer 100Hz. You can force
> top show satisfactory values if You set 2.6.XX
> system
> timer to 100, default is 1000
> include/asm/param.h:5:# define HZ               1000
> 
> check out procps package updates :o)
> 

But I don't use top: wrote my own code: get it from
/proc/<pid>/stat
Is that data bad too?

Hugo




		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - 250MB free storage. Do more. Manage less. 
http://info.mail.yahoo.com/mail_250

Re: Candidate multi-console friendly 2.6 scheduling?

[James v. Z. <ja...@dv...>]

Even using the gene algorithm library, the (zaphod) scheduler is not
radically changed, but rather it's operational parameters are constantly
evolving to suit the envioronment. So while it will adapt to suit
unusual loads or whatever is currently running, it's limitations remain
the same.

Such algorithms can be applied to any scheduler with adjustable "knobs",
and indeed wherever they are productive.

I think we need to seek several real-world benchmarks to compare the two
and be sure we are all looking at is valid. Using a phillips based video
capture card with 2.6.10, with the additional overhead of advanced
motion detection based deinterlacing, consumes 18 to 19% CPU. (TVTime)
That's a 2.8Athlon. About the same for playing back an avi with Xine...

I should be booting RH9 to directly compare...

Is it a case of 2.4 shows 20% on Xine, 2.6=30%
or 2.4=20% 2.6=70% ?

Suggested change noted, I'll have to check for the x86-64 fix too, I
think it might be there... -shall apply soon hopefully with some fresher
alsa code too :-)

J



On Wed, 2005-01-12 at 22:40, Hugo Vanwoerkom wrote:
> --- Aivils Stoss <ai...@la...> wrote:
> 
> > top LIES! top must be "tuned" according system
> > timer.
> > under 2.4.XX system timer 100Hz. You can force
> > top show satisfactory values if You set 2.6.XX
> > system
> > timer to 100, default is 1000
> > include/asm/param.h:5:# define HZ               1000
> > 
> > check out procps package updates :o)
> > 
> 
> But I don't use top: wrote my own code: get it from
> /proc/<pid>/stat
> Is that data bad too?
> 
> Hugo
> 
> 
> 
> 
> 		
> __________________________________ 
> Do you Yahoo!? 
> Yahoo! Mail - 250MB free storage. Do more. Manage less. 
> http://info.mail.yahoo.com/mail_250
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by: Beat the post-holiday blues
> Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
> It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> Linuxconsole-dev mailing list
> Lin...@li...
> https://lists.sourceforge.net/lists/listinfo/linuxconsole-dev
> 

Re: Candidate multi-console friendly 2.6 scheduling?

[Hugo V. <hvw...@ya...>]

--- Hugo Vanwoerkom <hvw...@ya...> wrote:

> 
> --- Aivils Stoss <ai...@la...> wrote:
> 
> > top LIES! top must be "tuned" according system
> > timer.
> > under 2.4.XX system timer 100Hz. You can force
> > top show satisfactory values if You set 2.6.XX
> > system
> > timer to 100, default is 1000
> > include/asm/param.h:5:# define HZ              
> 1000

Was set to 100 already.
H.



> > 
> > check out procps package updates :o)
> > 
> 
> But I don't use top: wrote my own code: get it from
> /proc/<pid>/stat
> Is that data bad too?
> 
> Hugo
> 
> 
> 
> 
> 		
> __________________________________ 
> Do you Yahoo!? 
> Yahoo! Mail - 250MB free storage. Do more. Manage
> less. 
> http://info.mail.yahoo.com/mail_250
> 
> 
>
-------------------------------------------------------
> The SF.Net email is sponsored by: Beat the
> post-holiday blues
> Get a FREE limited edition SourceForge.net t-shirt
> from ThinkGeek.
> It's fun and FREE -- well,
> almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> Linuxconsole-dev mailing list
> Lin...@li...
>
https://lists.sourceforge.net/lists/listinfo/linuxconsole-dev
> 



		
__________________________________ 
Do you Yahoo!? 
Take Yahoo! Mail with you! Get it on your mobile phone. 
http://mobile.yahoo.com/maildemo 

Re: Candidate multi-console friendly 2.6 scheduling?

[Helge H. <hel...@ai...>]

On Wed, Jan 12, 2005 at 04:40:05AM -0800, Hugo Vanwoerkom wrote:
> 
> But I don't use top: wrote my own code: get it from
> /proc/<pid>/stat
> Is that data bad too?
> 
Accuracy varies a lot between 2.4 and 2.6, due to the changed timer.
The worst case is processes that do periodically work with
short sleeps in between.  

If you want a good idea about how much "idle" time there is, do this:
1. Run your apps at normal priority for some time.
2. Run a cpu hog, something like  " for (i=0;;) ++i;" at the lowest
   possible priority.   Then check to see how much "work" the cpu
   hog manages to do (how far did it count, watch out for wraparound)
   in a given amount of time while the other stuff was running.

If the cpu hog gets less works done under 2.6, then you can say that
2.6 has more overhead (or at least a different take on
priority-based scheduling).  

Running the important program at realtime priority wil take care of even the
scheduling differences, but requires root privileges.

Helge Hafting 

Re: Candidate multi-console friendly 2.6 scheduling?

[James v. Z. <ja...@dv...>]

> Running the important program at realtime priority wil take care of even the
> scheduling differences, but requires root privileges.

ruby.vz kernel includes realtime LSM which allows non-root users to
request realtime priviledges. select default capabilities as a module
and realtime capabilities will appear as an option.
possibly real security risk. PlanetCCRMA includes sys-v init tools for
handling load/unload of realtime module, realtime capabilities can be
spcified by user-ID or Any.

For me this is necessary for compatibility with jack audio connection
kit and lowest latency audio performance.

Config_Hz is also available in kernel config

J


On Thu, 2005-01-13 at 09:16, Helge Hafting wrote:
> On Wed, Jan 12, 2005 at 04:40:05AM -0800, Hugo Vanwoerkom wrote:
> > 
> > But I don't use top: wrote my own code: get it from
> > /proc/<pid>/stat
> > Is that data bad too?
> > 
> Accuracy varies a lot between 2.4 and 2.6, due to the changed timer.
> The worst case is processes that do periodically work with
> short sleeps in between.  
> 
> If you want a good idea about how much "idle" time there is, do this:
> 1. Run your apps at normal priority for some time.
> 2. Run a cpu hog, something like  " for (i=0;;) ++i;" at the lowest
>    possible priority.   Then check to see how much "work" the cpu
>    hog manages to do (how far did it count, watch out for wraparound)
>    in a given amount of time while the other stuff was running.
> 
> If the cpu hog gets less works done under 2.6, then you can say that
> 2.6 has more overhead (or at least a different take on
> priority-based scheduling).  
> 
> Running the important program at realtime priority wil take care of even the
> scheduling differences, but requires root privileges.
> 
> Helge Hafting 
> 

Re: Candidate multi-console friendly 2.6 scheduling?

[Aivils S. <ai...@la...>]

On Thu, 13 Jan 2005, James van Zeeland wrote:

> 
> > Running the important program at realtime priority wil take care of even the
> > scheduling differences, but requires root privileges.

In my opinions, real time scheduling must be used delicate in
multiuser systems. Another users may got extreme long response.
CPU is single. Another user processes and users too sleep until
real time task runs :)

Aivils