unofficial linux-ck repo with cpu optimized packages is ONLINE

ummakynes · 2012-09-25 19:06:29

P4 64bit seems to be working fine, thanks bunches!

graysky · 2012-09-25 19:37:12

ummakynes wrote:

P4 64bit seems to be working fine, thanks bunches!

You're welcome. Enjoy.

rudiger · 2012-09-26 07:04:09

Thanks Graysky for your work. The linux-ck package is wonderfull.

IsSuE · 2012-09-26 08:31:10

rudiger wrote:

Thanks Graysky for your work. The linux-ck package is wonderfull.

Fully agreed!
Been running your packages for quite some time now, never had any serious trouble. Keep up your good work, it is very appreciated

graysky · 2012-09-26 09:05:57

Thanks for the kind works, guys. Feels good to give back to the Arch Community.

SiD · 2012-09-26 13:35:55

IsSuE wrote:

rudiger wrote:
Thanks Graysky for your work. The linux-ck package is wonderfull.
Fully agreed!
Been running your packages for quite some time now, never had any serious trouble. Keep up your good work, it is very appreciated

+1

Jristz · 2012-09-26 15:25:40

@Graysky I Know that I'm refering to a more technical point, any diference in perfomance or new/old cpu-flags-command or any not mentioned in the wikia

graysky · 2012-09-26 19:02:33

Jristz wrote:

@Graysky I Know that I'm refering to a more technical point, any diference in perfomance or new/old cpu-flags-command or any not mentioned in the wikia

Not sure what you mean... the nconfig offers only so many CPU-specific options. As noted, the kernel source overrides any settings you place in CFLAGS (in /etc/makepkg.conf).

Šaran · 2012-09-28 08:20:22

Just wanted to say thanks! I'm using corex for few days, and there is almost no difference in responsiveness under high load. Right now I'm building Cyanogenmod at 8.56 load and only fans are telling that laptop is under heavy load. With vanilla kernel browsing while building was almost impossible.

SanskritFritz · 2012-09-28 10:09:26

Šaran wrote:

Just wanted to say thanks! I'm using corex for few days, and there is almost no difference in responsiveness under high load. Right now I'm building Cyanogenmod at 8.56 load and only fans are telling that laptop is under heavy load. With vanilla kernel browsing while building was almost impossible.

Also enable BFQ and copy a big file to an usb stick

SiD · 2012-09-28 18:59:49

SanskritFritz wrote:

Also enable BFQ and copy a big file to an usb stick

Hmm, just tested to copy a 3.8GB file to a stick

CFQ: 5,45min
BFQ: 6:16min

Same file, same stick (file deleted from stick before second copy)

But it's offtopc, sorry.

SanskritFritz · 2012-09-29 11:17:16

SiD wrote:

SanskritFritz wrote:
Also enable BFQ and copy a big file to an usb stick
Hmm, just tested to copy a 3.8GB file to a stick
CFQ: 5,45min
BFQ: 6:16min
Same file, same stick (file deleted from stick before second copy)
But it's offtopc, sorry.

That is the whole point of BFQ. It is slower, but the system remains responsive. I remember in the past how my whole desktop froze for minutes when I had to copy files in the background.
Sorry for the offtopic.

PostBlue · 2012-09-29 11:22:38

I didn't find the relevant information in this 40-pages thread (I may haven't seen it), but is there an easy way to get "b43-firmware-ck" instead of/besides the broadcom-wl-ck ? It would be welcome.

PS : thx for this work, it's working awesomely on both my desktops.

graysky · 2012-09-29 11:41:56

To my knowledge, there isn't such a package. You can easily modify the non ck version in the aur to work either by deleting the depends array or modifying it to require linux-ck.

http://aur.archlinux.org/packages.php?ID=21690

PostBlue · 2012-09-29 15:06:52

Ok, I'll test it like that.

mikronimo · 2012-10-04 07:02:19

I, sorry for my posting in AUR; upgrade gone well; thanks.

graysky · 2012-10-04 08:14:14

mikronimo wrote:

I, sorry for my posting in AUR; upgrade gone well; thanks.

Glad it is working for you. There is no need to be sorry

domac · 2012-10-08 22:25:48

Linux-ck is awesome! Thanks for sharing!

graysky · 2012-10-10 15:48:57

EDIT: I spoke too soon. I cannot build this package with the unofficial patchset + the ARCH config files. My testing was indeed built with the base build, but I enabled the localmodconfig setting and thus only built a subset of all the modules (unique to my system). When I disabled this option and built the full package, I get build errors preventing compilation. As a result, we have to wait for a fix. Back to 3.5.6.

Here are the warnings from the build log so the problem lies somewhere with these options:

% grep warn what_the_fuck
cc1: some warnings being treated as errors
cc1: some warnings being treated as errors
drivers/atm/ambassador.c:1971:19: warning: ‘fw’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/base/regmap/regmap.c:1172:6: warning: ‘ret’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/gpu/drm/nouveau/nv40_pm.c:164:41: warning: ‘log2P’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/gpu/drm/nouveau/nv40_pm.c:165:38: warning: ‘M2’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/gpu/drm/nouveau/nv40_pm.c:165:45: warning: ‘M1’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/gpu/drm/nouveau/nv40_pm.c:165:25: warning: ‘N2’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/gpu/drm/nouveau/nv40_pm.c:165:51: warning: ‘N1’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/media/common/tuners/tda18212.c:309:2: warning: ‘val’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/media/common/tuners/tda18218.c:310:2: warning: ‘val’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/message/i2o/i2o_config.c:888:19: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/message/i2o/i2o_config.c:943:10: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/mfd/tps65217.c:169:13: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
include/linux/mtd/map.h:331:11: warning: ‘r.x[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
include/linux/mtd/map.h:331:11: warning: ‘r.x[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
include/linux/mtd/map.h:331:11: warning: ‘r.x[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
include/linux/mtd/map.h:331:11: warning: ‘r.x[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
include/linux/mtd/map.h:331:11: warning: ‘r.x[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/net/ethernet/amd/nmclan_cs.c:625:3: warning: ‘pcmcia_request_exclusive_irq’ is deprecated (declared at include/pcmcia/ds.h:201) [-Wdeprecated-declarations]
drivers/media/dvb/frontends/tda8261_cfg.h:68:21: warning: ‘t_state.bandwidth’ may be used uninitialized in this function [-Wuninitialized]
drivers/rtc/rtc-m41t80.c:216:12: warning: ‘m41t80_rtc_alarm_irq_enable’ defined but not used [-Wunused-function]
drivers/rtc/rtc-m41t80.c:238:12: warning: ‘m41t80_rtc_set_alarm’ defined but not used [-Wunused-function]
drivers/rtc/rtc-m41t80.c:308:12: warning: ‘m41t80_rtc_read_alarm’ defined but not used [-Wunused-function]
drivers/scsi/aic94xx/aic94xx_sds.c:597:21: warning: ‘offs’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/staging/bcm/CmHost.c:1401:39: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/staging/bcm/CmHost.c:1440:37: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/staging/bcm/CmHost.c:1454:35: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/net/ethernet/i825xx/znet.c:868:29: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
drivers/net/ethernet/i825xx/znet.c:875:29: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
drivers/scsi/mpt2sas/mpt2sas_scsih.c:5811:40: warning: unused variable ‘event_data’ [-Wunused-variable]
drivers/media/video/m5mols/m5mols.h:326:15: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/media/video/m5mols/m5mols.h:331:9: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
drivers/media/video/m5mols/m5mols.h:326:15: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/media/video/m5mols/m5mols.h:331:9: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
drivers/media/video/m5mols/m5mols_controls.c:466:2: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
drivers/media/video/m5mols/m5mols.h:326:15: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
drivers/media/video/m5mols/m5mols.h:331:9: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
drivers/staging/comedi/drivers/vmk80xx.c:920:37: warning: ‘reg[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/staging/comedi/drivers/vmk80xx.c:566:28: warning: ‘reg[0]’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/scsi/mvsas/mv_sas.c:1156:34: warning: comparison between ‘enum sas_device_type’ and ‘enum sas_dev_type’ [-Wenum-compare]
drivers/scsi/mvsas/mv_sas.c:1159:39: warning: comparison between ‘enum sas_device_type’ and ‘enum sas_dev_type’ [-Wenum-compare]
drivers/scsi/qla2xxx/qla_gs.c:1339:7: warning: array subscript is above array bounds [-Warray-bounds]
drivers/scsi/qla2xxx/qla_gs.c:1663:15: warning: array subscript is above array bounds [-Warray-bounds]
drivers/staging/media/lirc/lirc_serial.c:1301:1: warning: return from incompatible pointer type [enabled by default]
drivers/net/wireless/ipw2x00/libipw_wx.c:526:21: warning: unused variable ‘dev’ [-Wunused-variable]
drivers/net/ethernet/realtek/atp.c:871:26: warning: ‘mc_filter[0]’ may be used uninitialized in this function [-Wuninitialized]
drivers/staging/slicoss/slicoss.c:1493:19: warning: operation on ‘rspq->pageindex’ may be undefined [-Wsequence-point]
drivers/scsi/advansys.c:71:2: warning: #warning this driver is still not properly converted to the DMA API [-Wcpp]
drivers/staging/sm7xxfb/sm7xxfb.c:112:19: warning: ‘sm7xx_vga_setup’ defined but not used [-Wunused-function]
drivers/virtio/virtio_mmio.c:526:4: warning: format ‘%u’ expects argument of type ‘unsigned int *’, but argument 4 has type ‘resource_size_t *’ [-Wformat]
drivers/staging/wlan-ng/p80211netdev.c:355:24: warning: ‘p80211_wep.data’ may be used uninitialized in this function [-Wuninitialized]
drivers/net/wireless/wl3501_cs.c:655:9: warning: ‘sig.status’ is used uninitialized in this function [-Wuninitialized]
drivers/net/wireless/wl3501_cs.c:658:12: warning: ‘sig.cap_info’ may be used uninitialized in this function [-Wuninitialized]
drivers/net/wireless/wl3501_cs.c:660:12: warning: ‘sig.cap_info’ may be used uninitialized in this function [-Wuninitialized]
drivers/net/wireless/wl3501_cs.c:667:46: warning: ‘sig.ssid.el.len’ may be used uninitialized in this function [-Wuninitialized]

ORIGINAL POST:
tl;dr - I pushed the unofficial ck1 into the linux-ck package and am rebuilding packages for the repo now. I have a high confidence in the unofficial patch and will use it unless you guys tell me otherwise with issues if you experience them.

Details - A tragedy that has befallen the Kolivas family, which has focused CK's attentions elsewhere. Our wishes of warmth and strength go out to the Kolivas family! An active member of the community, Martin, has stepped up to fill the temporary void by porting both BFS v0.242 and CK1 from the 3.5 tree into the 3.6 tree.

I have done some initially benchmarking on Martin's port and have found it to preform very well in my make benchmark detailed below -- on par with other official CK releases. These data coupled with the broad adoption of these unofficial patches into Oleksandr Natalenko's linux-pf patchset make me feel confident in including these in linux-ck both in the AUR and in repo-ck.

Benchmark results:

The -ck patched kernel clearly differentiates itself from mainline with n=28 runs doing a `make -j8 bzImage` on v3.6.1.

Link to my benchmark script on github.

Details:
1) It is a non-latency based measure.
2) Compilation benchmark using gcc to “make bzImage” for a preconfigured linux 3.6.1 build.
3) Runs benchmarks 28 times totally to get a decent number of observations for a statistical comparison.
4) Results are how many seconds it took to compile on a 3370K @ 4.5 GHz.
5) Make is run with 8 threads (4 physical cores and 4 HT cores).

Last edited by graysky (2012-10-10 16:46:29)

Jristz · 2012-10-10 16:06:58

When I ask this (you new post) IS that I want read

Is clear that bfq is good and this is the clearest brenchmark,
But you say about the pf-patch that you gona to include it?? when??? or I read wrong???

graysky · 2012-10-10 16:38:03

@Jr - No. My point in mentioning linux-pf is that it another major patchset with a substantial user base has adpoted the unofficial patchset. That is all.

Do not confuse bfs with bfq. They are different. The bfq is only for disk IO; the bfs is for CPU. Both bot these are included in linux-ck.

utops · 2012-10-10 17:57:50

Thx for your work.
Edit
one question:
bfq is on via elevator=bfq, but how to enable bfs?

Last edited by utops (2012-10-10 18:01:36)

graysky · 2012-10-10 18:05:22

utops wrote:

bfq is on via elevator=bfq, but how to enable bfs?

BFS is hardcoded into linux-ck. You can check it in your dmesg like so:

% dmesg | grep BFS   
[    0.578275] BFS CPU scheduler v0.424 by Con Kolivas.

Last edited by graysky (2012-10-11 02:04:56)

graysky · 2012-10-13 13:39:31

OK! CK just released an OFFICIAL CK1 with bfs v0.425 that patches into the 3.6 tree, however it has a slight problem and will not build with guest virtualization enabled which as you know is part of the stock ARCH config. I emailed him and hope to have a fix soon. I will keep you updated.

graysky · 2012-10-13 16:37:41

OK! Problems solved and 3.6.2-1 is building for the repo now. Check it in a few hours.

Arch Linux

#976 2012-09-25 19:06:29

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#977 2012-09-25 19:37:12

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#978 2012-09-26 07:04:09

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#979 2012-09-26 08:31:10

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#980 2012-09-26 09:05:57

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#981 2012-09-26 13:35:55

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#982 2012-09-26 15:25:40

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#983 2012-09-26 19:02:33

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#984 2012-09-28 08:20:22

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#985 2012-09-28 10:09:26

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#986 2012-09-28 18:59:49

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#987 2012-09-29 11:17:16

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#988 2012-09-29 11:22:38

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#989 2012-09-29 11:41:56

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#990 2012-09-29 15:06:52

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#991 2012-10-04 07:02:19

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#992 2012-10-04 08:14:14

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#993 2012-10-08 22:25:48

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#994 2012-10-10 15:48:57

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#995 2012-10-10 16:06:58

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#996 2012-10-10 16:38:03

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#997 2012-10-10 17:57:50

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#998 2012-10-10 18:05:22

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#999 2012-10-13 13:39:31

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

#1000 2012-10-13 16:37:41

Re: unofficial linux-ck repo with cpu optimized packages is ONLINE

Board footer