Dell Inspiron 6000

[VinzC]

Try recompiling both Totem/Xine with CFLAGS="-march=pentium3 -pipe -O2 -fomit-frame-pointer". Could you also post the flags you USEd to compile Totem/Xine? I suppose you've also switched to GCC 3.4? You could also switch back to GCC 3.3 and recompile both packages. I doubt the kernel has anything to do with it but I'm not a specialist (yet ).

[juniper]

currently, my CFLAGS are

-march=pentium3 -O2 -pipe

will the fomit-frame-pointer make a difference? what is it?

j

[pblinux]

DVD playback still stutters. Any working solutions yet?

Average ~2.4MB/sec read from DVD

Code: Select all

>hdparm -Tt /dev/dvd
/dev/dvd:
 Timing cached reads:   1416 MB in  2.00 seconds = 707.96 MB/sec
 Timing buffered disk reads:    8 MB in  3.36 seconds =   2.38 MB/sec

libata.atapi_enabled=1 in kernel line

Code: Select all

>grep kernel /boot/grub/grub.conf
        kernel /boot/linux-2.6.15-suspend2-r6/bzImage root=/dev/sda3 resume2=/dev/sda2 video=vesafb:ywrap,mtrr,1024x768-16 CONSOLE=/dev/tty1 rootflags=commit=300 libata.atapi_enabled=1

ATA_ENABLE_ATAPI defined in libata.h

Code: Select all

>grep def.*ATA.*ENABLE /usr/src/linux/include/linux/libata.h
#undef ATA_ENABLE_PATA          /* define to enable PATA support in some
#define ATAPI_ENABLE_DMADIR     /* enables ATAPI DMADIR bridge support */
#define ATA_ENABLE_ATAPI        /* define to enable ATAPI support */

Kernel .config

Code: Select all

>grep ATA /usr/src/linux/.config
CONFIG_X86_MCE_NONFATAL=y
# CONFIG_ATALK is not set
# CONFIG_ATA_OVER_ETH is not set
# ATA/ATAPI/MFM/RLL support
CONFIG_SCSI_SATA=y
# CONFIG_SCSI_SATA_AHCI is not set
# CONFIG_SCSI_SATA_SVW is not set
CONFIG_SCSI_ATA_PIIX=y
# CONFIG_SCSI_SATA_MV is not set
# CONFIG_SCSI_SATA_NV is not set
# CONFIG_SCSI_SATA_QSTOR is not set
# CONFIG_SCSI_SATA_PROMISE is not set
# CONFIG_SCSI_SATA_SX4 is not set
# CONFIG_SCSI_SATA_SIL is not set
# CONFIG_SCSI_SATA_SIL24 is not set
# CONFIG_SCSI_SATA_SIS is not set
# CONFIG_SCSI_SATA_ULI is not set
# CONFIG_SCSI_SATA_VIA is not set
# CONFIG_SCSI_SATA_VITESSE is not set
CONFIG_SCSI_SATA_INTEL_COMBINED=y
# CONFIG_SCSI_EATA is not set
# CONFIG_USB_STORAGE_DATAFAB is not set

[juniper]

VinzC,

actually, i have gcc-3.3.6. perhaps i should upgrade. i am emerging totem/xine now with the cflags you suggested (will upgrade gcc after).

j

[VinzC]

VinzC,

actually, i have gcc-3.3.6. perhaps i should upgrade. i am emerging totem/xine now with the cflags you suggested (will upgrade gcc after).

j

Actually I thought you had the same flags as I do (-march=pentium-m -mtune=pentium-m -O3 -fomit-frame-pointer). I basically suggested you to downgrade optimizations but it was true only if your flags were different. As Sejam pointed out in his initial post, you should switch to GCC 3.4 and use its specific flags.

As far as the kernel is concerned, it is best to switch to the 2.6.14 series as these include the libata patch. All you have to do is add kernel argument libata.atapi_enabled=1. And if you switch to 2.6.16 series, ieee80211 and ipw2200 will be built-in (or as modules), no more need to emerge from portage.

It is really worth the try. Now you'll have many packages to recompile, it's true. First try Xorg and Gnome (or KDE). Then recompile system packages if it worked. If you want to rush I think it is now safe to recompile every single package at once using GCC 3.4. Just follow GCC Migration Guide.

Frame pointers are special stack decoration of function calls to allow backtracing I think. Remove them for better performance. You won't need more optimization as GCC is smart enough to auto-select which ones to apply.

[raf_kimmig]

hey there, just one question:

is there anybody out there who actually got suspend to ram working reliably with fglrx?

right now im running 2.6.15.6 (patchset from http://rtr.ca/dell_i9300/ ) xorg 6.8 and version 8.22.5 of ati's proprietary

suspend to ram works most of the time but sometimes the screen just stays black when trying to resume and so far
i am unable to safely reproduce that issue - but it occurs to often to ignore it...

ive got simply no idea what to do now, forget vbetool, switches to console aso - i've written so many suspend/resume scripts
and i thought being successful so many times, and than boom - next resume : guess the color of my screen

what im going to do next is - having a look at the dsdt although, iirc, it did only have a minor number of common warnings and no errors at all

i'd greatly appreciate any hints on what to do, where to look or whatever you think might help me

regards

raf

[hiboo]

well, I have also tried so many combinaison of configuration.... always failing...
I regret to have bought a laptop with an ATI...
I haven't tried, and I think it is impossible but let's have hope : can I change my graphic card ?

[seppelrockt]

DVD playback still stutters. Any working solutions yet?

Average ~2.4MB/sec read from DVD

libata.atapi_enabled=1 in kernel line

Code: Select all

>grep kernel /boot/grub/grub.conf
        kernel /boot/linux-2.6.15-suspend2-r6/bzImage root=/dev/sda3 resume2=/dev/sda2 video=vesafb:ywrap,mtrr,1024x768-16 CONSOLE=/dev/tty1 rootflags=commit=300 libata.atapi_enabled=1

ATA_ENABLE_ATAPI defined in libata.h

Yes I know this thread is very long but please try to search it for infos anyway - repeating the same things over and over again makes the problem even worse. The solution is simple: grep a kernel >= 2.6.15 (vanilla works, gentoo sources not tested from my side) and DO NOT PATCH/MODIFY the kernel anymore (this was for 2.6.14 and older ONLY). Just libata.atapi_enabled=1 as kernel line should do the trick.

And yes, we are comming closer to my rewrite of this Inspiron 6000 guide Maybe in four weeks or so I will finaly do it. Does anybody have some knowledge source about different GCC versions (3.4.5 vs. 4.x) and there impacts on the speed of pentium-m systems? I just wonder whether it might be worth using GCC-4.x when it hits ~x86 or whether I should stay on good ol' GCC 3.4.5 ...

[hiboo]

Yes I know this thread is very long but please try to search it for infos anyway - repeating the same things over and over again makes the problem even worse. The solution is simple: grep a kernel >= 2.6.15 (vanilla works, gentoo sources not tested from my side)

works with the stable gentoo sources for me

[geekounet]

Does anybody have some knowledge source about different GCC versions (3.4.5 vs. 4.x) and there impacts on the speed of pentium-m systems? I just wonder whether it might be worth using GCC-4.x when it hits ~x86 or whether I should stay on good ol' GCC 3.4.5 ...

I'm using GCC 4.1 without any problem
GCC 4.0 doesn't bring much than GCC 3.4, only the internal functionment has changed, but results of compilations are the same.
But GCC 4.1 compiles in a different way of the previous (i don't know much more about this), and C++ code compiles faster.
I can't say if moving to GCC 4.1 bring better performance because i changed my CFLAGS at the same moment, but since these changes my system is effectively more performant.
My CFLAGS :

CFLAGS="-O3 -march=pentium-m -pipe -fomit-frame-pointer -mmmx -msse -msse2 -mfpmath=sse,387 -fforce-addr"
CXXFLAGS="${CFLAGS} -fvisibility-inlines-hidden"

Please can you tell me if the RAM on this laptop is DDR or DDR2 ? I'm just wondering if I didn't wrote a mistake in my signature ? ^^

[raf_kimmig]

if i remove my mem theres a warning that says : "use ddr2 only"

[geekounet]

Ok, thanks

[pblinux2]

Yes I know this thread is very long but please try to search it for infos anyway - repeating the same things over and over again makes the problem even worse. The solution is simple: grep a kernel >= 2.6.15 (vanilla works, gentoo sources not tested from my side) and DO NOT PATCH/MODIFY the kernel anymore (this was for 2.6.14 and older ONLY). Just libata.atapi_enabled=1 as kernel line should do the trick.

I'm a little frustrated here...I printed out every page of this thread and went through each post one-by-one. Some of the posts are outdated, and some are just incorrect, and it's pretty hard to do a search for this problem when it's described differently be every person (choppy playback, stuttering movies, dvd problems, atapi config, etc). When I posted my message, it wasn't even possible to perform a search on a specific thread, and I was wading through several hundred "solutions" that were irrelevant to an Inspiron 6000.

Cut the new guy a little slack, eh?

For starters, what speeds should I be getting from an "hdparm -Tt /dev/dvd" if things are configured correctly.
Secondly, from [post=3151836]my post[/post], what do I have configured incorrectly?

Currently, doing a "hdparm -Tt /dev/dvd" just returns:
read(2097152) returned 1179648 bytes
Timing buffered disk reads: read() failed: Input/output error

What am I missing here?[/url]

[VinzC]

For starters, what speeds should I be getting from an "hdparm -Tt /dev/dvd" if things are configured correctly.
Secondly, from [post=3151836]my post[/post], what do I have configured incorrectly?

Currently, doing a "hdparm -Tt /dev/dvd" just returns:
read(2097152) returned 1179648 bytes
Timing buffered disk reads: read() failed: Input/output error

What am I missing here?[/url]

When I had my Inspiron 6000, I never - since 2.6.15 series - had to do anything to the kernel. Even Gentoo Sources did work properly. Be careful *not* to select any ATA stuff, like Sejam initially posted in this thread.

Now hdparm won't work if you don't select any ATA/... in your kernel. Kernel argument libata.atapi_enabled=1 is enough to enable DMA on the SATA (IDE) disk *plus* the CDROM drive.

I've used Vanilla Sources 2.6.15 and 2.6.16 series (patched only with Spock's fbsplash patchset) and Gentoo Sources 2.6.15 series. Both give the same results with reading DVD: no hitch, no freeze, no problem. Playing DVDs was quite normal and fluid - video files as well.

So if you have hdparm tell you something else than "wrong operation" or "operation not permitted", then you've enabled ATA in your kernel. hdparm should not report anything else. Just dmesg | grep -i dma to check whether DMA is enabled on both your hard disk and CDROM/DVD drive.

[mrbox]

I'm having some problems with X300 and dual head.
When I use the ati drivers (latest) without a secondary display I don’t get any error’s.
But when I hookup a secondary display (xinerama, different resolutions), the items displayed at the opposed display show up as garbled lines at the other display.
Any ideas ?

[SOLVED] If you enable DRM then it doesn't show

[seppelrockt]

I'm a little frustrated here ... [...] Cut the new guy a little slack, eh?

Sorry if that turned out wrong - I didn't blame you to be lazy or a noobie that doens't use the search function. And I know quite good how frustrating it is to search this post (I usally use the search function of my browser with highlighting keywords).

If you had a look at the whole post once and didn't find your information or are not able to destingish the outdated and the up to date infos then you shall recieve an answer to your question here - well, and you got it, right? But most people do not read a 20+ pages thread that might be related to there question and put this question right to the end of the ever repeating chain of questions.

So please try to understand my little frustration, too - cause I a) answered the libata + kernel-2.6.15 question many times here in the forum (and this thread, too), b) I use this thread as a source of information just like you and the more often we repeat the same questions and answers the more long and confusing the whole thing gonna get and c) I have serious plans to rewrite the whole thing - so it hurts to see the thread getting longer and longer without new information. Concluding, I just wanted to express in my last post that the infos you are looking for are allready there - I didn't looked at the number of your posts or something ...

P.S. Volunteers for the rewrite thing that want to help are welcome - drop me a PM or something. I am looking for people that have suspend modes working and people with x300 ati graphics in particular.

[econan]

I am looking for people that have suspend modes working and people with x300 ati graphics in particular.

I have suspend working 90% and I have x300 ATI.
Didn't do something major. I have a very very basic suspend script.
Let me know how i can help...

[seppelrockt]

I am looking for people that have suspend modes working and people with x300 ati graphics in particular.

I have suspend working 90% and I have x300 ATI.
Didn't do something major. I have a very very basic suspend script.
Let me know how i can help...

My rough plan is to do a fresh step by step install of Gentoo 2006.0 on this laptop and write down every step it needs - this is pretty much about bringing the infos from this thread in a good order When it comes e.g. to the step "configuring the xserver" I can of cause only write this part for the graphic card I have (Intel). So I will leave a gap there for X300 graphics first and hope somebody else will step in and send me the infos for that graphic card. Then I will add this infos in my post. That how I think about it. Don't know when I will find the time to write it actually.

[amaroc]

... to see the thread getting longer and longer without new information.
...
I am looking for people that have suspend modes working and people with x300 ati graphics in particular.

OK, I own the i915 version, but some news regarding suspend
The recently released kernel 2.6.16 makes suspend-to-ram more reliable. I had suspend-to-disk already stable w/ 2.6.15, but suspend-to-ram didn't resume some times (about the 20th...30th try) due to SATA issues. I have now tested 2.6.16 about 50 times - no issues so far.
It's vanilla 2.6.16 + patchset from http://rtr.ca/dell_i9300/kernel/kernel-2.6.16/ + vesa-tng from http://dev.gentoo.org/~spock/projects/v ... 6.16.patch

Maybe it helps for the ati-version as well.

btw: I'm really happy with this laptop and would suggest it for use w/ linux and Gentoo in particular.

[torchZ06]

anyone running a 2.6.16 vanilla kernel on their 6000? i put it on mine and it no longer has support for the SCSI_QLA2200 that's listed as a requiremet for the CD/DVD and of course it broke mine as a result-- had to go back to a 2.6.15-gentoo and it started working again.

[amaroc]

anyone running a 2.6.16 vanilla kernel on their 6000? i put it on mine and it no longer has support for the SCSI_QLA2200 that's listed as a requiremet for the CD/DVD and of course it broke mine as a result-- had to go back to a 2.6.15-gentoo and it started working again.

As written earlier, I'm running a vanilla 2.6.16 w/ some patches from http://rtr.ca/dell_i9300/kernel/kernel-2.6.16/. DVD playback and burnig works fine and suspend-to-ram is really stable.
I don't have the QLA2200 in 2.6.16 and I didn't had it enabled in 2.6.15. I don't know why it should be enabled.
Did you had problems during compiling the kernel? Then you should have run

Code: Select all

make oldconfig

before running

Code: Select all

make menuconfig
make && make modules_install

If it's really a run-time problem I would recommend the patches from above.

[econan]

As written earlier, I'm running a vanilla 2.6.16 w/ some patches from http://rtr.ca/dell_i9300/kernel/kernel-2.6.16/. DVD playback and burnig works fine and suspend-to-ram is really stable.

I second that! (with ATI)

[VinzC]

As written earlier, I'm running a vanilla 2.6.16 w/ some patches from http://rtr.ca/dell_i9300/kernel/kernel-2.6.16/. DVD playback and burnig works fine and suspend-to-ram is really stable.

I'll especially try the SD Card Reader patches on my 9400 for I noticed things are a little bit different with 2.6.16 kernels.

EDIT: works !

[seppelrockt]

As written earlier, I'm running a vanilla 2.6.16 w/ some patches from http://rtr.ca/dell_i9300/kernel/kernel-2.6.16/. DVD playback and burnig works fine and suspend-to-ram is really stable.

I'll especially try the SD Card Reader patches on my 9400 for I noticed things are a little bit different with 2.6.16 kernels.

EDIT: works !

Good news - the card reader stuff made it into the future 2.6.17 kernel - so once a rc1 is out grab it and put one more patch in the trash bin

[VinzC]

Good news - the card reader stuff made it into the future 2.6.17 kernel - so once a rc1 is out grab it and put one more patch in the trash bin

Yup, I know .