[Qemu-devel] AHCI Port Interrupt Enable register cleaning on soft reset

[Qemu-devel] AHCI Port Interrupt Enable register cleaning on soft reset

[Alexander Motin]

Hi.

I've found that FreeBSD AHCI driver doesn't work with AHCI hardware
emulation of QEMU 0.15.0. I believe the problem is on QEMU's side. As I
see, it clears port's Interrupt Enable register each time when reset of
any level happens. Is is reasonable for the global controller reset. It
is probably not good, but acceptable for FreeBSD driver for the port
hard reset. But it is IMO wrong for the device soft reset. None of real
hardware I know behaves that way.

This patch fixes the problem for me:
http://people.freebsd.org/~mav/qemu.ahci.patch

-- 
Alexander Motin

Re: [Qemu-devel] AHCI Port Interrupt Enable register cleaning on soft reset

[Alexander Graf]

Am 11.09.2011 um 16:43 schrieb Alexander Motin <mav@FreeBSD.org>:

> Hi.
> 
> I've found that FreeBSD AHCI driver doesn't work with AHCI hardware
> emulation of QEMU 0.15.0. I believe the problem is on QEMU's side. As I
> see, it clears port's Interrupt Enable register each time when reset of
> any level happens. Is is reasonable for the global controller reset. It
> is probably not good, but acceptable for FreeBSD driver for the port
> hard reset. But it is IMO wrong for the device soft reset. None of real
> hardware I know behaves that way.
> 
> This patch fixes the problem for me:
> http://people.freebsd.org/~mav/qemu.ahci.patch

Ah, cool! So FreeBSD works with AHCI using this patch? Please send it again as an inline patch (if really really hard not 100% important) and add a signed-off-by line (very important) to the patch.


Alex

Re: [Qemu-devel] AHCI Port Interrupt Enable register cleaning on soft reset

[Alexander Motin]

Alexander Graf wrote:
> Am 11.09.2011 um 16:43 schrieb Alexander Motin <mav@FreeBSD.org>:
>> I've found that FreeBSD AHCI driver doesn't work with AHCI hardware
>> emulation of QEMU 0.15.0. I believe the problem is on QEMU's side. As I
>> see, it clears port's Interrupt Enable register each time when reset of
>> any level happens. Is is reasonable for the global controller reset. It
>> is probably not good, but acceptable for FreeBSD driver for the port
>> hard reset. But it is IMO wrong for the device soft reset. None of real
>> hardware I know behaves that way.
>>
>> This patch fixes the problem for me:
>> http://people.freebsd.org/~mav/qemu.ahci.patch
> 
> Ah, cool! So FreeBSD works with AHCI using this patch? 

Yes. I haven't done deep testing to guarantee there is no other issues,
but at least disk is properly detected now.

> Please send it again as an inline patch (if really really hard not 100% important) and add a signed-off-by line (very important) to the patch.

OK. Here it is:

Signed-off-by: Alexander Motin <mav@FreeBSD.org>

--- hw/ide/ahci.c.prev	2011-09-11 16:39:53.000000000 +0300
+++ hw/ide/ahci.c	2011-09-11 16:39:48.000000000 +0300
@@ -505,10 +505,7 @@ static void ahci_reset_port(AHCIState *s
     ide_bus_reset(&d->port);
     ide_state->ncq_queues = AHCI_MAX_CMDS;

-    pr->irq_stat = 0;
-    pr->irq_mask = 0;
     pr->scr_stat = 0;
-    pr->scr_ctl = 0;
     pr->scr_err = 0;
     pr->scr_act = 0;
     d->busy_slot = -1;
@@ -1157,12 +1154,17 @@ void ahci_uninit(AHCIState *s)
 void ahci_reset(void *opaque)
 {
     struct AHCIPCIState *d = opaque;
+    AHCIPortRegs *pr;
     int i;

     d->ahci.control_regs.irqstatus = 0;
     d->ahci.control_regs.ghc = 0;

     for (i = 0; i < d->ahci.ports; i++) {
+        pr = &d->ahci.dev[i].port_regs;
+        pr->irq_stat = 0;
+        pr->irq_mask = 0;
+        pr->scr_ctl = 0;
         ahci_reset_port(&d->ahci, i);
     }
 }

-- 
Alexander Motin

Re: [Qemu-devel] AHCI Port Interrupt Enable register cleaning on soft reset

[Kevin Wolf]

Am 12.09.2011 10:19, schrieb Alexander Motin:
> Alexander Graf wrote:
>> Am 11.09.2011 um 16:43 schrieb Alexander Motin <mav@FreeBSD.org>:
>>> I've found that FreeBSD AHCI driver doesn't work with AHCI hardware
>>> emulation of QEMU 0.15.0. I believe the problem is on QEMU's side. As I
>>> see, it clears port's Interrupt Enable register each time when reset of
>>> any level happens. Is is reasonable for the global controller reset. It
>>> is probably not good, but acceptable for FreeBSD driver for the port
>>> hard reset. But it is IMO wrong for the device soft reset. None of real
>>> hardware I know behaves that way.
>>>
>>> This patch fixes the problem for me:
>>> http://people.freebsd.org/~mav/qemu.ahci.patch
>>
>> Ah, cool! So FreeBSD works with AHCI using this patch? 
> 
> Yes. I haven't done deep testing to guarantee there is no other issues,
> but at least disk is properly detected now.
> 
>> Please send it again as an inline patch (if really really hard not 100% important) and add a signed-off-by line (very important) to the patch.
> 
> OK. Here it is:
> 
> Signed-off-by: Alexander Motin <mav@FreeBSD.org>
> 
> --- hw/ide/ahci.c.prev	2011-09-11 16:39:53.000000000 +0300
> +++ hw/ide/ahci.c	2011-09-11 16:39:48.000000000 +0300
> @@ -505,10 +505,7 @@ static void ahci_reset_port(AHCIState *s
>      ide_bus_reset(&d->port);
>      ide_state->ncq_queues = AHCI_MAX_CMDS;
> 
> -    pr->irq_stat = 0;
> -    pr->irq_mask = 0;
>      pr->scr_stat = 0;
> -    pr->scr_ctl = 0;
>      pr->scr_err = 0;
>      pr->scr_act = 0;
>      d->busy_slot = -1;
> @@ -1157,12 +1154,17 @@ void ahci_uninit(AHCIState *s)
>  void ahci_reset(void *opaque)
>  {
>      struct AHCIPCIState *d = opaque;
> +    AHCIPortRegs *pr;
>      int i;
> 
>      d->ahci.control_regs.irqstatus = 0;
>      d->ahci.control_regs.ghc = 0;
> 
>      for (i = 0; i < d->ahci.ports; i++) {
> +        pr = &d->ahci.dev[i].port_regs;
> +        pr->irq_stat = 0;
> +        pr->irq_mask = 0;
> +        pr->scr_ctl = 0;
>          ahci_reset_port(&d->ahci, i);
>      }
>  }
> 

Thanks, applied to the block branch.

Kevin