[SOLVED] Marvell 88SE6111 IDE Problems

Message

slaterson · Post by **slaterson** » Wed Aug 06, 2008 8:59 pm

removing the jumper did the trick. no performance gain but its nice to see that the system is operating as it should and the configuration is correct.

thanks!

Monogram · Post by **Monogram** » Wed Aug 27, 2008 5:20 am

I just wanted to give my thanks... great job: emerald... I was trying to resolve this issue for over a week now without much luck - till I came across this posting. I applied all patches and now I got my optical drive on Marvell 6121 controller working fine.

refer to: http://forums.gentoo.org/viewtopic-t-705323.html

May I inquire as to how you achieved this? / how you edited / patched the drivers? I'd love to know.

Thanks again.

SysGhost · Post by **SysGhost** » Fri Sep 12, 2008 10:50 pm

Greetings.

Nice to see there is a solution for the by-now-well-known problem with ahci claiming all SATA controllers it can find.

I tried to reach this patch http://www.home.liji.de/kernel/tmp/ahci ... hips.patch for a while now. Either you are not up and running very much, or it has gone missing.

I could hold a copy of it on my own web-server that are up 24/7 for better availability.

emerald · Post by **emerald** » Sat Sep 13, 2008 11:40 am

SysGhost wrote:I tried to reach this patch http://www.home.liji.de/kernel/tmp/ahci ... hips.patch for a while now. Either you are not up and running very much, or it has gone missing.

I was off for 2 days for hardware change, now back on

Anyway, i keep the patches in another location too, just not as current is in the former one.

2.6.25
http://www.emi-sama.de/patches/kernel/2 ... 61xx.patch
http://www.emi-sama.de/patches/kernel/2 ... hips.patch

2.6.26
http://www.emi-sama.de/patches/kernel/2 ... 61xx.patch
http://www.emi-sama.de/patches/kernel/2 ... hips.patch

2.6.27
http://www.emi-sama.de/patches/kernel/2 ... 61xx.patch
http://www.emi-sama.de/patches/kernel/2 ... hips.patch

SysGhost · Post by **SysGhost** » Sat Sep 13, 2008 4:31 pm

Thanks! I'll see if these solves my problems too...

As I mentioned. I'll keep a copy of those patches available aswell: (should be available 24 / 7 as long the server don't explode and burn up violently)
(Those who do download: I don't have my mimetype configured ... yet, so rightclick and choose "save link as". Or cut'n'paste into a simple textdocument)

kernel 2.6.25:
http://ghostnet.sytes.net/patches/kerne ... 61xx.patch
http://ghostnet.sytes.net/patches/kerne ... hips.patch

kernel 2.6.26:
http://ghostnet.sytes.net/patches/kerne ... 61xx.patch
http://ghostnet.sytes.net/patches/kerne ... hips.patch

Monogram · Post by **Monogram** » Fri Sep 26, 2008 2:46 am

I seem to be having a problem (initially: http://forums.gentoo.org/viewtopic-t-705323.html).

I've got everything working thanks to emerald little patch as far as my SATA optical drive on the Marvell 6121 and bios raid on the ati sb600 controller.

However on the Marvel 6111 (which I believe is in fact an IDE interface) I have a ide/pata hard-drive that seems to be going through ahci drivers?

I have Marvell drivers compiled as module as well as related stuff.

I've noticed that the overall operation on the drive is rather slow (~ 2megs a sec)... and I cant seem to be able to enable DMA on the drive.

Any suggestions?

emerald · Post by **emerald** » Fri Sep 26, 2008 8:59 am

Monogram wrote: However on the Marvel 6111 (which I believe is in fact an IDE interface) I have a ide/pata hard-drive that seems to be going through ahci drivers?

I have Marvell drivers compiled as module as well as related stuff.

Read this post http://forums.gentoo.org/viewtopic-p-51 ... ml#5166961 about the AHCI and Marvell driver interaction.

Edit: wrong post

SysGhost · Post by **SysGhost** » Thu Oct 02, 2008 8:50 pm

Update on my progress here:

That simple "AHCI don't claim 61xx" patch worked just fine.
Next step was trying to get that 6121 driver compiled and loaded (I prefer having it in modular form)

For a while I was looking under SATA section of the menu-configuration (make menuconfig)
...nothing!

After a while I got myself to look under scsi section and there I found it:

Device Drivers --->
SCSI device support --->
SCSI low-level drivers --->
Marvell Storage Controller 6121/6122/6141/6145

Compiled, installed and loaded, I got it working... almost...

As the driver tries to detect my SATA Bluray reader ( Asus BC-1205PT ) it keeps spitting out:
Interrupt Error: 0x40000000 orgIntStatus: 0x40000001 completeSlot=0x1

Still unable to use the BD drive of mine, but under windows it works jyust fine, so I guess there are no hardware faults. It's just the driver not supporting this device... yet...

Guess I have to wait til the next version of the driver is released...

...oh! By the way... where do I download the driver?
Marvell themselves seems to have only network card drivers and Asus have nothing but a preconfigured 2.6.12 kernel.

emerald · Post by **emerald** » Thu Oct 02, 2008 10:51 pm

SysGhost wrote:Update on my progress here:
[...]
As the driver tries to detect my SATA Bluray reader ( Asus BC-1205PT ) it keeps spitting out:
Interrupt Error: 0x40000000 orgIntStatus: 0x40000001 completeSlot=0x1

Still unable to use the BD drive of mine, but under windows it works jyust fine, so I guess there are no hardware faults. It's just the driver not supporting this device... yet...

Guess I have to wait til the next version of the driver is released...

Why don't you just connect the BD drive to the Intel sata ports and put move a harddisk, which seems to be working fine, to the marvell chip (the harddrive might be a bit slower but you got a working BD drive instead, for me there'd be no more question)?

SysGhost wrote: ...oh! By the way... where do I download the driver?
Marvell themselves seems to have only network card drivers and Asus have nothing but a preconfigured 2.6.12 kernel.

The drivers i found at the Asus support webpage for my mainboard, those drivers were declared working at 2.6.18, so i just made them working with the newer kernels. Whether there will ever be a newer driver from Marvell i dare not guess. Marvell themselves don't support non-networking chips directly, so don't expect to find any on their webpage.

SysGhost · Post by **SysGhost** » Fri Oct 03, 2008 2:05 pm

Good idea, but in my case it's the SB600 Chipset since I'm on one "Asus M3A32 MVP Deluxe"
Problems remains, module mv61xx still causes:

"Interrupt Error: 0x40000000 orgIntStatus: 0x40000001 completeSlot=0x1."

It doesn't matter if there's something connected or not.
I start to believe it's a hardware conflict, maybe a simple IRQ conflict... (Strange that it's works just fine under windows, but Windows maybe handle the IRQ's differently)

EDIT: letting the SATA AHCI driver take control of the Marvell 6121 chipset seems to work flawlessly, except the fact I cannot use any Optical drives on it (most SATA optical drives don't support AHCI) ... strange in my opinion...

EDIT 2:
Letting SATA AHCI control Marvell 6121 with Asus BC-1205PT Bluray reader attached causes these errors:

Code: Select all

ata5.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
ata5.00: cmd a0/00:00:00:02:00/00:00:00:00:00/a0 tag 0 pio 16388 in
         cdb 51 00 00 00 00 00 00 00  02 00 00 00 00 00 00 00
         res 40/00:03:00:00:00/00:03:00:00:00/a0 Emask 0x4 (timeout)
ata5.00: status: { DRDY }
ata5: hard resetting link
ata5: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata5.00: qc timeout (cmd 0xa1)
ata5.00: failed to IDENTIFY (I/O error, err_mask=0x4)
ata5.00: revalidation failed (errno=-5)
ata5: failed to recover some devices, retrying in 5 secs
ata5: hard resetting link
ata5: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata5.00: qc timeout (cmd 0xa1)
ata5.00: failed to IDENTIFY (I/O error, err_mask=0x4)
ata5.00: revalidation failed (errno=-5)
ata5: failed to recover some devices, retrying in 5 secs
ata5: hard resetting link
ata5: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata5.00: qc timeout (cmd 0xa1)
ata5.00: failed to IDENTIFY (I/O error, err_mask=0x4)
ata5.00: revalidation failed (errno=-5)
ata5.00: disabled
ata5: hard resetting link
ata5: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata5: EH complete

These errors are because the Bluray reader aren't AHCI compartible, and the controller is initiated in AHCI mode.
Any AHCI compartible device works just fine however.

blossa · Post by **blossa** » Sat Oct 11, 2008 8:24 am

I see that this thread is a few month old now... Do we still need to patch and so on or is this fixed any other way by now?

Thanks emerald for your work!

Dew · Post by **Dew** » Thu Nov 13, 2008 10:48 am

Yeah, is everything fine in 2.6.27-r2?
And, can somebody tell me, which patch do I need? And what to compile in kernel for it to work, of course.

KarlisRepsons · Post by **KarlisRepsons** » Wed Dec 03, 2008 11:01 am

Nice, there is at least a patch for this thing, I used both patches, but the problem still remains about eSATA port, which simply doesn't work.
How about others' experience with it?

emerald · Post by **emerald** » Wed Dec 03, 2008 11:15 am

For me the eSata ports work without problems using the marvell driver, not ahci.

KarlisRepsons · Post by **KarlisRepsons** » Wed Dec 03, 2008 11:17 am

Have you applied both patches?
Also I found, that somehow IDE becomes unusable, if tuxonice kernel is where it is patched.
I have:

in lsmod: "mv61xx 69288 0",
in kernel messages:
Marvell (S)ATA Controller is found, using IRQ 16.
scsi6 : Marvell 88SE61xx Storage Controller

It was allright with gentoo-sources, but not with tuxonice. Any idea? I think, I made no mistake...

KarlisRepsons · Post by **KarlisRepsons** » Sat Dec 06, 2008 3:35 pm

Well, I don't understand how can that be: I can start booting gentoo livecd (then it fails, no driver), so hardware is operational, but somehow IDE DVDrw ceased to work even with previously working kernel. Some kind of nightmare. If you have no immediate idea, you can safely neglect...

emerald · Post by **emerald** » Sat Dec 06, 2008 3:54 pm

I don't use TuxOnIce since for everything I need the in-kernel suspend/hibernate support is more than adequate for me.
Since the Marvell driver is not incorporated in the live cd the ide port will only work in legacy mode if it even works at all.
So to use a ide cd/dvd drive you'll have to build your own livecd (or change the kernel of the one you wish to use) or get something else to boot off.
Wit your own kernel it should work, if ont just have a look at previous posts (possibly in different threads), there were many things explained and discussed in detail already.
If you wish to use TuxOnIce i'd still recommend to try a regular kernel first, if it works then to try TuxOnIce but there you'll probably have to try to fix most of your problems by yourself, possibly by patching the driver. If you get it to work well with patches i'll gladly incorporate those into the current patches.

KarlisRepsons · Post by **KarlisRepsons** » Sat Dec 06, 2008 4:30 pm

No, no, I understand about livecd.
The scary thing is: I was using this driver with gentoo-sources-2.6.25-r7 and it worked. Now, after some kernel recompiles and BIOS resetting (jumper to delete what is set in BIOS) and reconfiguring (well, I tried to get the same, but not sure) it ceased to work even with same kernel.
Hardware problem? Well, how could it partially boot CD then?

Simply... stupid. "What has happened," I can only think.

KarlisRepsons · Post by **KarlisRepsons** » Sat Dec 06, 2008 6:25 pm

So I apply just one of patches for marvell driver to gentoo-sources, device itself is working (could boot with), but the rest is failure.
Maybe some idea where I could ask how to debug?

Please take a look at your dmesg, what is different from this:

Marvell (S)ATA Controller is found, using IRQ 16.
success to init chip
scsi0 : Marvell 88SE61xx Storage Controller
Driver 'sd' needs updating - please use bus_type methods
Driver 'sr' needs updating - please use bus_type methods
ahci 0000:00:1f.2: version 3.0
ACPI: PCI Interrupt 0000:00:1f.2 -> GSI 19 (level, low) -> IRQ 19
ahci 0000:00:1f.2: AHCI 0001.0200 32 slots 6 ports 3 Gbps 0x3f impl SATA mode
ahci 0000:00:1f.2: flags: 64bit ncq sntf stag pm led clo pmp pio slum part
PCI: Setting latency timer of device 0000:00:1f.2 to 64
scsi1 : ahci
scsi2 : ahci
scsi3 : ahci
scsi4 : ahci
scsi5 : ahci
scsi6 : ahci
ata1: SATA max UDMA/133 abar m2048@0xf8ffe800 port 0xf8ffe900 irq 379
ata2: SATA max UDMA/133 abar m2048@0xf8ffe800 port 0xf8ffe980 irq 379
ata3: SATA max UDMA/133 abar m2048@0xf8ffe800 port 0xf8ffea00 irq 379
ata4: SATA max UDMA/133 abar m2048@0xf8ffe800 port 0xf8ffea80 irq 379
ata5: SATA max UDMA/133 abar m2048@0xf8ffe800 port 0xf8ffeb00 irq 379
ata6: SATA max UDMA/133 abar m2048@0xf8ffe800 port 0xf8ffeb80 irq 379
ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata1.00: ATA-8: WDC WD6401AALS-00L3B2, 01.03B01, max UDMA/133
ata1.00: 1250263728 sectors, multi 0: LBA48 NCQ (depth 31/32)
ata1.00: configured for UDMA/133
ata2: SATA link down (SStatus 0 SControl 300)
ata3: SATA link down (SStatus 0 SControl 300)
ata4: SATA link down (SStatus 0 SControl 300)
ata5: SATA link down (SStatus 0 SControl 300)
ata6: SATA link down (SStatus 0 SControl 300)
scsi 1:0:0:0: Direct-Access ATA WDC WD6401AALS-0 01.0 PQ: 0 ANSI: 5
sd 1:0:0:0: [sda] 1250263728 512-byte hardware sectors (640135 MB)

emerald · Post by **emerald** » Sat Dec 06, 2008 6:43 pm

Code: Select all

mv_thor 0000:03:00.0: setting latency timer to 64
Marvell (S)ATA Controller is found, using IRQ 16.
 for non-raid devices
success to init chip
scsi1 : Marvell 88SE61xx Storage Controller
Driver 'sd' needs updating - please use bus_type methods
scsi 1:0:20:0: CD-ROM            HL-DT-ST DVD-RAM GSA-H22L 1.02 PQ: 0 ANSI: 0
Driver 'sr' needs updating - please use bus_type methods
ahci 0000:00:1f.2: version 3.0
ahci 0000:00:1f.2: PCI INT B -> GSI 22 (level, low) -> IRQ 22
ahci 0000:00:1f.2: AHCI 0001.0200 32 slots 6 ports 3 Gbps 0x3f impl SATA mode
ahci 0000:00:1f.2: flags: 64bit ncq sntf stag pm led clo pmp pio slum part ems
ahci 0000:00:1f.2: setting latency timer to 64
scsi2 : ahci
scsi3 : ahci
scsi4 : ahci
scsi5 : ahci
scsi6 : ahci
scsi7 : ahci

For me it looks like that, the marvell chip is initialized in non-raid mode (which i should be correct for dvdrom drives).

KarlisRepsons · Post by **KarlisRepsons** » Mon Dec 08, 2008 9:13 am

Just for fun, this is what asus support answers:
1) Could you install fresh linux?
2)The motherboard officially supports Windows XP/Vista operating systems.
So, could you please install a XP OS so that we can figure out this issue easily?

slaterson · Post by **slaterson** » Fri Jun 12, 2009 12:44 am

hi emerald. i tried your website from a few posts above but it seems to be down. is there a patch for kernel 2.6.29? i'd like to upgrade but, of course, need the marvell support.

thanks!

emerald · Post by **emerald** » Fri Jun 12, 2009 1:53 pm

The current kernel contains all necessary drivers, the ahci driver and the marvell pata driver will support those chips now, so the marvell driver is no longer necessary.

slaterson · Post by **slaterson** » Fri Jun 12, 2009 3:30 pm

excellent, thanks. i took a quick look last night but didn't see it, i'll have a closer look today.

slaterson · Post by **slaterson** » Fri Jun 12, 2009 6:31 pm

i went through the .29 kernel this morning but couldn't locate the mv_thor driver. has it been combined with the mv driver? there is something very different with the new .29 kernel, as when i first booted i got a kernel panic. had to reboot with the old kernel and change my hard drive device names from sdb to sda. i remember when i first used you patch after applying the patches i had to rename from sda to sdb, so it seems i am missing something or things have changed under the covers.

thanks