hp dv8000t dead on resume from RAM

hp dv8000t dead on resume from RAM

[Brannon Barrett Klopfer]

Howdy,

My hp dv8000t (core duo) is completely dead on resume from RAM (no caps
lock, sysrq, netconsole,
nothing). I've tried a recent (2.6.18-rc4) kernel running almost entirely
naked, so to speak (~990K) -- no support for:

SMP
preempt
modules
networking (+ enet drivers, unless running netconsole)
USB and SATA (not at same time; rootfs is either usb drive or SATA [ext2/3])
FireWire
cpufreq
framebuffer (vga=0)
audio
PCMCIA
IDE (for cdrom)

I've tried both native SATA (ahci) and legacy (ata_piix), but same result
w/both -- completely dead on resume from RAM. Blindly entering commands
does nothing, and running "$suspend ; $shutdown" does nothing either. I've
also tried with and without noapic, and a number of other kernel
paramaters, but nothing seems to work.

Be more than happy to try out patches, etc. to get this thing working.
Additionally, if someone could point me to that "beep on resume" patch,
that'd be great.

Some machine/debugging info (cpuinfo, messages, lspci, etc.) at:

http://www.stanford.edu/~bklopfer/wontsuspend/

Thanks,
Brannon Klopfer

Re: hp dv8000t dead on resume from RAM

[Rafael J. Wysocki]

On Tuesday 08 August 2006 19:15, Brannon Barrett Klopfer wrote:
> Howdy,
> 
> My hp dv8000t (core duo) is completely dead on resume from RAM (no caps
> lock, sysrq, netconsole,
> nothing). I've tried a recent (2.6.18-rc4) kernel running almost entirely
> naked, so to speak (~990K) -- no support for:
> 
> SMP
> preempt
> modules
> networking (+ enet drivers, unless running netconsole)
> USB and SATA (not at same time; rootfs is either usb drive or SATA [ext2/3])
> FireWire
> cpufreq
> framebuffer (vga=0)
> audio
> PCMCIA
> IDE (for cdrom)
> 
> I've tried both native SATA (ahci) and legacy (ata_piix), but same result
> w/both -- completely dead on resume from RAM. Blindly entering commands
> does nothing, and running "$suspend ; $shutdown" does nothing either. I've
> also tried with and without noapic, and a number of other kernel
> paramaters, but nothing seems to work.
> 
> Be more than happy to try out patches, etc. to get this thing working.
> Additionally, if someone could point me to that "beep on resume" patch,
> that'd be great.

First, please apply the appended patch and try the following:

(1)
# echo testproc > /sys/power/disk
# echo disk > /sys/power/state

This should turn off the non-boot CPU, freeze all processes, wait for 5
seconds and then thaw the processes and the CPU.

(2)
# echo test > /sys/power/disk
# echo disk > /sys/power/state

This should turn off the non-boot CPU, freeze all processes, shrink
memory, suspend all devices, wait for 5 seconds, resume the devices etc.
IOW it does everything that's needed for a suspend except for actually
suspending.

Greetings,
Rafael

---
 include/linux/pm.h  |    4 +++-
 kernel/power/disk.c |   37 ++++++++++++++++++++++++++++---------
 2 files changed, 31 insertions(+), 10 deletions(-)

Index: linux-2.6.18-rc3-mm2/include/linux/pm.h
===================================================================
--- linux-2.6.18-rc3-mm2.orig/include/linux/pm.h
+++ linux-2.6.18-rc3-mm2/include/linux/pm.h
@@ -116,7 +116,9 @@ typedef int __bitwise suspend_disk_metho
 #define	PM_DISK_PLATFORM	((__force suspend_disk_method_t) 2)
 #define	PM_DISK_SHUTDOWN	((__force suspend_disk_method_t) 3)
 #define	PM_DISK_REBOOT		((__force suspend_disk_method_t) 4)
-#define	PM_DISK_MAX		((__force suspend_disk_method_t) 5)
+#define	PM_DISK_TEST		((__force suspend_disk_method_t) 5)
+#define	PM_DISK_TESTPROC	((__force suspend_disk_method_t) 6)
+#define	PM_DISK_MAX		((__force suspend_disk_method_t) 7)
 
 struct pm_ops {
 	suspend_disk_method_t pm_disk_mode;
Index: linux-2.6.18-rc3-mm2/kernel/power/disk.c
===================================================================
--- linux-2.6.18-rc3-mm2.orig/kernel/power/disk.c
+++ linux-2.6.18-rc3-mm2/kernel/power/disk.c
@@ -70,7 +70,7 @@ static inline void platform_finish(void)
 
 static int prepare_processes(void)
 {
-	int error;
+	int error = 0;
 
 	pm_prepare_console();
 
@@ -83,6 +83,12 @@ static int prepare_processes(void)
 		goto thaw;
 	}
 
+	if (pm_disk_mode == PM_DISK_TESTPROC) {
+		printk("swsusp debug: Waiting for 5 seconds.\n");
+		mdelay(5000);
+		goto thaw;
+	}
+
 	/* Free memory before shutting down devices. */
 	if (!(error = swsusp_shrink_memory()))
 		return 0;
@@ -119,11 +125,19 @@ int pm_suspend_disk(void)
 	if (error)
 		return error;
 
+	if (pm_disk_mode == PM_DISK_TESTPROC)
+		goto Thaw;
+
 	error = device_suspend(PMSG_FREEZE);
 	if (error) {
 		printk("Some devices failed to suspend\n");
-		unprepare_processes();
-		return error;
+		goto Thaw;
+	}
+
+	if (pm_disk_mode == PM_DISK_TEST) {
+		printk("swsusp debug: Waiting for 5 seconds.\n");
+		mdelay(5000);
+		goto Done;
 	}
 
 	pr_debug("PM: snapshotting memory.\n");
@@ -139,15 +153,16 @@ int pm_suspend_disk(void)
 			power_down(pm_disk_mode);
 		else {
 			swsusp_free();
-			unprepare_processes();
-			return error;
+			goto Thaw;
 		}
-	} else
+	} else {
 		pr_debug("PM: Image restored successfully.\n");
+	}
 
 	swsusp_free();
  Done:
 	device_resume();
+ Thaw:
 	unprepare_processes();
 	return error;
 }
@@ -241,6 +256,8 @@ static const char * const pm_disk_modes[
 	[PM_DISK_PLATFORM]	= "platform",
 	[PM_DISK_SHUTDOWN]	= "shutdown",
 	[PM_DISK_REBOOT]	= "reboot",
+	[PM_DISK_TEST]		= "test",
+	[PM_DISK_TESTPROC]	= "testproc",
 };
 
 /**
@@ -295,17 +312,19 @@ static ssize_t disk_store(struct subsyst
 		}
 	}
 	if (mode) {
-		if (mode == PM_DISK_SHUTDOWN || mode == PM_DISK_REBOOT)
+		if (mode == PM_DISK_SHUTDOWN || mode == PM_DISK_REBOOT ||
+		     mode == PM_DISK_TEST || mode == PM_DISK_TESTPROC) {
 			pm_disk_mode = mode;
-		else {
+		} else {
 			if (pm_ops && pm_ops->enter &&
 			    (mode == pm_ops->pm_disk_mode))
 				pm_disk_mode = mode;
 			else
 				error = -EINVAL;
 		}
-	} else
+	} else {
 		error = -EINVAL;
+	}
 
 	pr_debug("PM: suspend-to-disk mode set to '%s'\n",
 		 pm_disk_modes[mode]);

Re: hp dv8000t dead on resume from RAM

[Brannon Klopfer]

Rafael J. Wysocki wrote:
> On Tuesday 08 August 2006 19:15, Brannon Barrett Klopfer wrote:
>   
>> Howdy,
>>
>> My hp dv8000t (core duo) is completely dead on resume from RAM (no caps
>> lock, sysrq, netconsole,
>> nothing). I've tried a recent (2.6.18-rc4) kernel running almost entirely
>> naked, so to speak (~990K) -- no support for:
>>
>> SMP
>> preempt
>> modules
>> networking (+ enet drivers, unless running netconsole)
>> USB and SATA (not at same time; rootfs is either usb drive or SATA [ext2/3])
>> FireWire
>> cpufreq
>> framebuffer (vga=0)
>> audio
>> PCMCIA
>> IDE (for cdrom)
>>
>> I've tried both native SATA (ahci) and legacy (ata_piix), but same result
>> w/both -- completely dead on resume from RAM. Blindly entering commands
>> does nothing, and running "$suspend ; $shutdown" does nothing either. I've
>> also tried with and without noapic, and a number of other kernel
>> paramaters, but nothing seems to work.
>>
>> Be more than happy to try out patches, etc. to get this thing working.
>> Additionally, if someone could point me to that "beep on resume" patch,
>> that'd be great.
>>     
>
> First, please apply the appended patch and try the following:
>
> (1)
> # echo testproc > /sys/power/disk
> # echo disk > /sys/power/state
>
> This should turn off the non-boot CPU, freeze all processes, wait for 5
> seconds and then thaw the processes and the CPU.
>   
Works fine. FWIW, after applying the patchs (so as to have 
2.6.18-rc3-mm2), my internal keyboard didn't work, so I used a USB one. 
It could be my simple .config'ing error, didn't spend much time with it, 
but know that I did use a USB keyboard, hence USB support in the kernel.
> (2)
> # echo test > /sys/power/disk
> # echo disk > /sys/power/state
>
> This should turn off the non-boot CPU, freeze all processes, shrink
> memory, suspend all devices, wait for 5 seconds, resume the devices etc.
> IOW it does everything that's needed for a suspend except for actually
> suspending.
>   
With *legacy* ata_piix, it works fine, as does a "real" suspend-to-disk. 
It still *will not* resume properly from suspend-to-RAM, with and 
without noapic. FWIW, somehow the built-in keyboard managed to get its 
caps lock light on (I couldn't get it off again), and it *did* come back 
on when the machine resumed from RAM, though nothing else (network, usb, 
etc.) worked.

Using native ahci and the "test" suspend-to-disk, the system hangs. I 
did this from init=/bin/bash with vga=794 (so I could see all output), 
however, at least the last few lines match w/vga=0 (i.e., fb didn't 
affect problem). The hand-copied (pardon any typos) output is:
---
Disabling non-boot CPUs ...
CPU 1 is now offline
SMP alternatives: switching to UP code
CPU1 is down
Stopping tasks: =======|
Shrinking memory... done (0 pages freed)
Suspending device usbdev4.5
Suspending device usbdev4p5_ep82
Suspending device 4-2.1:1.1
Suspending device usbdev4.5_ep81
Suspending device 4-2.1:1.0
Suspending device usbdev4.5_ep00
Suspending device 4-2.1
Suspending device usbdev4.4
Suspending device usbdev4.4_ep81
Suspending device 4-2:1.0
Suspending device usbdev4.4_ep00
Suspending device 4-2
Suspending device vcsa1
Suspending device vcs1
Suspending device eisa.0
Suspending device i2c-0
Suspending device i2c-0
Suspending device psaux
Suspending device serio4
Suspending device serio3
Suspending device serio2
Suspending device serio1
Suspending device serio0
Suspending device i8042
# Short pause here...
atkbd.c: Spurious ACK on isa0060/serio0. Some program might be trying 
access hardware directly.
Suspending device usbdev5.1
Suspending device usbdev5.1_ep81
Suspending device 5-0:1.0
Suspending device usbdev5.1_ep00
Suspending device usb5
Suspending device usbdev4.1
Suspending device usbdev4.1_ep81
Suspending device 4-0:1.0
Suspending device usbdev4.1_ep00
Suspending device usb4
Suspending device usbdev3.1
Suspending device usbdev3.1_ep81
Suspending device 3-0:1.0
Suspending device usbdev3.1_ep00
Suspending device usb3
Suspending device usbdev2.1
Suspending device usbdev2.1_ep81
Suspending device 2-0:1.0
Suspending device usbdev2.1_ep00
Suspending device usb2
Suspending device usbdev1.1
Suspending device usbdev1.1_ep81
Suspending device 1-0:1.0
Suspending device usbdev1.1_ep00
Suspending device usb1
Suspending device 0:0:0:0
Suspending device target0:0:0
Suspending device host3
Suspending device host2
Suspending device host1
Suspending device host0
Suspending device 0.0
# Hangs here. Screen blanks in maybe 5-10min, won't turn back on; I 
assume this is the hardware doing something. Again, this is only with ahci.
---

I'll try to get that beep patch, to further diagnose things...

Thanks,
Brannon
> Greetings,
> Rafael
>
> ---
>  include/linux/pm.h  |    4 +++-
>  kernel/power/disk.c |   37 ++++++++++++++++++++++++++++---------
>  2 files changed, 31 insertions(+), 10 deletions(-)
<snip...patch applied cleanly>

Re: hp dv8000t dead on resume from RAM

[Rafael J. Wysocki]

On Wednesday 09 August 2006 20:01, Brannon Klopfer wrote:
> Rafael J. Wysocki wrote:
> > On Tuesday 08 August 2006 19:15, Brannon Barrett Klopfer wrote:
> >   
> >> Howdy,
> >>
> >> My hp dv8000t (core duo) is completely dead on resume from RAM (no caps
> >> lock, sysrq, netconsole,
> >> nothing). I've tried a recent (2.6.18-rc4) kernel running almost entirely
> >> naked, so to speak (~990K) -- no support for:
> >>
> >> SMP
> >> preempt
> >> modules
> >> networking (+ enet drivers, unless running netconsole)
> >> USB and SATA (not at same time; rootfs is either usb drive or SATA [ext2/3])
> >> FireWire
> >> cpufreq
> >> framebuffer (vga=0)
> >> audio
> >> PCMCIA
> >> IDE (for cdrom)
> >>
> >> I've tried both native SATA (ahci) and legacy (ata_piix), but same result
> >> w/both -- completely dead on resume from RAM. Blindly entering commands
> >> does nothing, and running "$suspend ; $shutdown" does nothing either. I've
> >> also tried with and without noapic, and a number of other kernel
> >> paramaters, but nothing seems to work.
> >>
> >> Be more than happy to try out patches, etc. to get this thing working.
> >> Additionally, if someone could point me to that "beep on resume" patch,
> >> that'd be great.
> >>     
> >
> > First, please apply the appended patch and try the following:
> >
> > (1)
> > # echo testproc > /sys/power/disk
> > # echo disk > /sys/power/state
> >
> > This should turn off the non-boot CPU, freeze all processes, wait for 5
> > seconds and then thaw the processes and the CPU.
> >   
> Works fine. FWIW, after applying the patchs (so as to have 
> 2.6.18-rc3-mm2), my internal keyboard didn't work, so I used a USB one. 

That is specific to 2.6.18-rc3-mm2 and has nothing to do with the suspend.
There should be a patch for that in hot-fixes.

> It could be my simple .config'ing error, didn't spend much time with it, 
> but know that I did use a USB keyboard, hence USB support in the kernel.
> > (2)
> > # echo test > /sys/power/disk
> > # echo disk > /sys/power/state
> >
> > This should turn off the non-boot CPU, freeze all processes, shrink
> > memory, suspend all devices, wait for 5 seconds, resume the devices etc.
> > IOW it does everything that's needed for a suspend except for actually
> > suspending.
> >   
> With *legacy* ata_piix, it works fine, as does a "real" suspend-to-disk. 
> It still *will not* resume properly from suspend-to-RAM, with and 
> without noapic.

I think I know what the problem is, but unfortunately I have no patch for
that.  Please look at
http://www.ussg.iu.edu/hypermail/linux/kernel/0607.3/1566.html
but the fix discussed there is for AMD/NVidia only.

> FWIW, somehow the built-in keyboard managed to get its  
> caps lock light on (I couldn't get it off again), and it *did* come back 
> on when the machine resumed from RAM, though nothing else (network, usb, 
> etc.) worked.
> 
> Using native ahci and the "test" suspend-to-disk, the system hangs. I 
> did this from init=/bin/bash with vga=794 (so I could see all output), 
> however, at least the last few lines match w/vga=0 (i.e., fb didn't 
> affect problem). The hand-copied (pardon any typos) output is:

Some additional patches are needed for the AHCI suspend, AFAIK.
I thought they were on the way to the mainline, but it looks like I was
wrong.  See: http://www.ussg.iu.edu/hypermail/linux/kernel/0607.2/0885.html

Greetings,
Rafael

Re: hp dv8000t dead on resume from RAM

[Brannon Barrett Klopfer]

Quoting "Rafael J. Wysocki" <rjw@sisk.pl>:

> On Wednesday 09 August 2006 20:01, Brannon Klopfer wrote:
> > Rafael J. Wysocki wrote:
> > > On Tuesday 08 August 2006 19:15, Brannon Barrett Klopfer wrote:
> > >
> > >> Howdy,
> > >>
> > >> My hp dv8000t (core duo) is completely dead on resume from RAM (no
> caps
> > >> lock, sysrq, netconsole,
> > >> nothing). I've tried a recent (2.6.18-rc4) kernel running almost
> entirely
> > >> naked, so to speak (~990K) -- no support for:
> > >>
> > >> SMP
> > >> preempt
> > >> modules
> > >> networking (+ enet drivers, unless running netconsole)
> > >> USB and SATA (not at same time; rootfs is either usb drive or SATA
> [ext2/3])
> > >> FireWire
> > >> cpufreq
> > >> framebuffer (vga=0)
> > >> audio
> > >> PCMCIA
> > >> IDE (for cdrom)
> > >>
> > >> I've tried both native SATA (ahci) and legacy (ata_piix), but same
> result
> > >> w/both -- completely dead on resume from RAM. Blindly entering
> commands
> > >> does nothing, and running "$suspend ; $shutdown" does nothing
> either. I've
> > >> also tried with and without noapic, and a number of other kernel
> > >> paramaters, but nothing seems to work.
> > >>
> > >> Be more than happy to try out patches, etc. to get this thing
> working.
> > >> Additionally, if someone could point me to that "beep on resume"
> patch,
> > >> that'd be great.
> > >>
> > >
> > > First, please apply the appended patch and try the following:
> > >
> > > (1)
> > > # echo testproc > /sys/power/disk
> > > # echo disk > /sys/power/state
> > >
> > > This should turn off the non-boot CPU, freeze all processes, wait for
> 5
> > > seconds and then thaw the processes and the CPU.
> > >
> > Works fine. FWIW, after applying the patchs (so as to have
> > 2.6.18-rc3-mm2), my internal keyboard didn't work, so I used a USB one.
>
> That is specific to 2.6.18-rc3-mm2 and has nothing to do with the
> suspend.
> There should be a patch for that in hot-fixes.
>
> > It could be my simple .config'ing error, didn't spend much time with
> it,
> > but know that I did use a USB keyboard, hence USB support in the
> kernel.
> > > (2)
> > > # echo test > /sys/power/disk
> > > # echo disk > /sys/power/state
> > >
> > > This should turn off the non-boot CPU, freeze all processes, shrink
> > > memory, suspend all devices, wait for 5 seconds, resume the devices
> etc.
> > > IOW it does everything that's needed for a suspend except for
> actually
> > > suspending.
> > >
> > With *legacy* ata_piix, it works fine, as does a "real"
> suspend-to-disk.
> > It still *will not* resume properly from suspend-to-RAM, with and
> > without noapic.
>
> I think I know what the problem is, but unfortunately I have no patch for
> that.  Please look at
> http://www.ussg.iu.edu/hypermail/linux/kernel/0607.3/1566.html
> but the fix discussed there is for AMD/NVidia only.

Hmmm...I did try a kernel with *no* SATA or IDE support, and used a pen
drive for my root; would that rule out that sort of problem, or does the
hardware freak out regardless of whether or not there's a driver for it?


> > Using native ahci and the "test" suspend-to-disk, the system hangs. I
> > did this from init=/bin/bash with vga=794 (so I could see all output),
> > however, at least the last few lines match w/vga=0 (i.e., fb didn't
> > affect problem). The hand-copied (pardon any typos) output is:
>
> Some additional patches are needed for the AHCI suspend, AFAIK.
> I thought they were on the way to the mainline, but it looks like I was
> wrong.  See:
> http://www.ussg.iu.edu/hypermail/linux/kernel/0607.2/0885.html

I tried to patch the kernel with the "test/testproc" suspend to disk trick,
but it didn't apply cleanly; I applied it cleanly to 2.6.18-rc4 and then
copied drivers/scsi/ahci.c to the "test/testproc" kernel, but it didn't
seem to make a difference.

Thanks,
Brannon

Re: hp dv8000t dead on resume from RAM

[Rafael J. Wysocki]

On Thursday 10 August 2006 20:29, Brannon Barrett Klopfer wrote:
> Quoting "Rafael J. Wysocki" <rjw@sisk.pl>:
> 
> > On Wednesday 09 August 2006 20:01, Brannon Klopfer wrote:
> > > Rafael J. Wysocki wrote:
> > > > On Tuesday 08 August 2006 19:15, Brannon Barrett Klopfer wrote:
> > > >
> > > >> Howdy,
> > > >>
> > > >> My hp dv8000t (core duo) is completely dead on resume from RAM (no
> > caps
> > > >> lock, sysrq, netconsole,
> > > >> nothing). I've tried a recent (2.6.18-rc4) kernel running almost
> > entirely
> > > >> naked, so to speak (~990K) -- no support for:
> > > >>
> > > >> SMP
> > > >> preempt
> > > >> modules
> > > >> networking (+ enet drivers, unless running netconsole)
> > > >> USB and SATA (not at same time; rootfs is either usb drive or SATA
> > [ext2/3])
> > > >> FireWire
> > > >> cpufreq
> > > >> framebuffer (vga=0)
> > > >> audio
> > > >> PCMCIA
> > > >> IDE (for cdrom)
> > > >>
> > > >> I've tried both native SATA (ahci) and legacy (ata_piix), but same
> > result
> > > >> w/both -- completely dead on resume from RAM. Blindly entering
> > commands
> > > >> does nothing, and running "$suspend ; $shutdown" does nothing
> > either. I've
> > > >> also tried with and without noapic, and a number of other kernel
> > > >> paramaters, but nothing seems to work.
> > > >>
> > > >> Be more than happy to try out patches, etc. to get this thing
> > working.
> > > >> Additionally, if someone could point me to that "beep on resume"
> > patch,
> > > >> that'd be great.
> > > >>
> > > >
> > > > First, please apply the appended patch and try the following:
> > > >
> > > > (1)
> > > > # echo testproc > /sys/power/disk
> > > > # echo disk > /sys/power/state
> > > >
> > > > This should turn off the non-boot CPU, freeze all processes, wait for
> > 5
> > > > seconds and then thaw the processes and the CPU.
> > > >
> > > Works fine. FWIW, after applying the patchs (so as to have
> > > 2.6.18-rc3-mm2), my internal keyboard didn't work, so I used a USB one.
> >
> > That is specific to 2.6.18-rc3-mm2 and has nothing to do with the
> > suspend.
> > There should be a patch for that in hot-fixes.
> >
> > > It could be my simple .config'ing error, didn't spend much time with
> > it,
> > > but know that I did use a USB keyboard, hence USB support in the
> > kernel.
> > > > (2)
> > > > # echo test > /sys/power/disk
> > > > # echo disk > /sys/power/state
> > > >
> > > > This should turn off the non-boot CPU, freeze all processes, shrink
> > > > memory, suspend all devices, wait for 5 seconds, resume the devices
> > etc.
> > > > IOW it does everything that's needed for a suspend except for
> > actually
> > > > suspending.
> > > >
> > > With *legacy* ata_piix, it works fine, as does a "real"
> > suspend-to-disk.
> > > It still *will not* resume properly from suspend-to-RAM, with and
> > > without noapic.
> >
> > I think I know what the problem is, but unfortunately I have no patch for
> > that.  Please look at
> > http://www.ussg.iu.edu/hypermail/linux/kernel/0607.3/1566.html
> > but the fix discussed there is for AMD/NVidia only.
> 
> Hmmm...I did try a kernel with *no* SATA or IDE support, and used a pen
> drive for my root; would that rule out that sort of problem, or does the
> hardware freak out regardless of whether or not there's a driver for it?

It may or may not, I can't tell really.

Please wait for the next -mm kernel to appear and try it (there should be
some generic IDE fixes in there).

> > > Using native ahci and the "test" suspend-to-disk, the system hangs. I
> > > did this from init=/bin/bash with vga=794 (so I could see all output),
> > > however, at least the last few lines match w/vga=0 (i.e., fb didn't
> > > affect problem). The hand-copied (pardon any typos) output is:
> >
> > Some additional patches are needed for the AHCI suspend, AFAIK.
> > I thought they were on the way to the mainline, but it looks like I was
> > wrong.  See:
> > http://www.ussg.iu.edu/hypermail/linux/kernel/0607.2/0885.html
> 
> I tried to patch the kernel with the "test/testproc" suspend to disk trick,
> but it didn't apply cleanly; I applied it cleanly to 2.6.18-rc4 and then
> copied drivers/scsi/ahci.c to the "test/testproc" kernel, but it didn't
> seem to make a difference.

For now, I have no other ideas.

Please create a bugzilla entry for your problem and describe what's wrong and
what you've done already to debug it.  Please also put some hardware and
configuration details in there (please add the address rjwysocki@sisk.pl
to the entry's Cc list).

Greetings,
Rafael