Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Meelis Roos]

(CC-s added)

> > > Tried 4.2.0-rc3-00246-g763e326 on A500 but it crashes on boot. This is 
> > > still present in todays 4.2.0-rc6+git. 4.1 was fine, I Will bisect but 
> > > it takes time.
> > > 
> > > PDC Stable Storage facility v0.30
> > > STI GSC/PCI core graphics driver Version 0.9b
> > > Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
> > > serial 0000:00:04.0: enabling device (0146 -> 0147)
> > > console [ttyS0] disabled
> > > 0000:00:04.0: ttyS0 at MMIO 0xfffffffff8000000 (irq = 21, base_baud = 115200) is a 16550A
> > > console [ttyS0] enabled
> > > console [ttyS0] enabled
> > > bootconsole [ttyB0] disabled
> > > bootconsole [ttyB0] disabled
> > > 0000:00:04.0: ttyS1 at MMIO 0xfffffffff8000008 (irq = 21, base_baud = 115200) is a 16550A
> > > 0000:00:04.0: ttyS2 at MMIO 0xfffffffff8000010 (irq = 21, base_baud = 115200) is a 16550A
> > > serial 0000:00:05.0: enabling device (0000 -> 0003)
> > > serial 0000:00:05.0: enabling SERR and PARITY (0003 -> 0143)
> > > 0000:00:05.0: ttyS3 at MMIO 0xfffffffff8003000 (irq = 22, base_baud = 115200) is a 16550A
> > > serial 0000:00:05.0: Couldn't register serial port 0, irq 22, type 2, error -28
> > > sym53c8xx 0000:00:01.0: enabling device (0000 -> 0003)
> > > sym53c8xx 0000:00:01.0: enabling SERR and PARITY (0003 -> 0143)
> > 
> > I'm seeing the same problem on my rp5470 with Kernel 4.2-rc7. My machine just hangs though and doesn't crash.
> 
> I did a bisect and commit 3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d ("PCI: Add pci_bus_addr_t") seems to be the culprit:
> http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d
> 
> Since the patch has "CC: stable@vger.kernel.org  # v3.19+", it might be the reason why you see kernel 3.19 failing too...

I just found the same commit breaking my A500 parisc machine by 
bisecting.

CC: patch author and linux-pci.

-- 
Meelis Roos (mroos@linux.ee)

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Helge Deller]

On 18.08.2015 20:44, Meelis Roos wrote:
> (CC-s added)
>
>>>> Tried 4.2.0-rc3-00246-g763e326 on A500 but it crashes on boot. This is
>>>> still present in todays 4.2.0-rc6+git. 4.1 was fine, I Will bisect but
>>>> it takes time.
>>>>
>>>> PDC Stable Storage facility v0.30
>>>> STI GSC/PCI core graphics driver Version 0.9b
>>>> Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
>>>> serial 0000:00:04.0: enabling device (0146 -> 0147)
>>>> console [ttyS0] disabled
>>>> 0000:00:04.0: ttyS0 at MMIO 0xfffffffff8000000 (irq = 21, base_baud = 115200) is a 16550A
>>>> console [ttyS0] enabled
>>>> console [ttyS0] enabled
>>>> bootconsole [ttyB0] disabled
>>>> bootconsole [ttyB0] disabled
>>>> 0000:00:04.0: ttyS1 at MMIO 0xfffffffff8000008 (irq = 21, base_baud = 115200) is a 16550A
>>>> 0000:00:04.0: ttyS2 at MMIO 0xfffffffff8000010 (irq = 21, base_baud = 115200) is a 16550A
>>>> serial 0000:00:05.0: enabling device (0000 -> 0003)
>>>> serial 0000:00:05.0: enabling SERR and PARITY (0003 -> 0143)
>>>> 0000:00:05.0: ttyS3 at MMIO 0xfffffffff8003000 (irq = 22, base_baud = 115200) is a 16550A
>>>> serial 0000:00:05.0: Couldn't register serial port 0, irq 22, type 2, error -28
>>>> sym53c8xx 0000:00:01.0: enabling device (0000 -> 0003)
>>>> sym53c8xx 0000:00:01.0: enabling SERR and PARITY (0003 -> 0143)
>>>
>>> I'm seeing the same problem on my rp5470 with Kernel 4.2-rc7. My machine just hangs though and doesn't crash.
>>
>> I did a bisect and commit 3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d ("PCI: Add pci_bus_addr_t") seems to be the culprit:
>> http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d
>>
>> Since the patch has "CC: stable@vger.kernel.org  # v3.19+", it might be the reason why you see kernel 3.19 failing too...
>
> I just found the same commit breaking my A500 parisc machine by
> bisecting.
>
> CC: patch author and linux-pci.

I think this is the problem:

--- a/drivers/pci/Kconfig
+++ b/drivers/pci/Kconfig
@@ -1,6 +1,10 @@
  #
  # PCI configuration
  #
+config PCI_BUS_ADDR_T_64BIT
+       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
+       depends on PCI


CONFIG_PCI_BUS_ADDR_T_64BIT gets now defined on all 64bit arches.
Then if CONFIG_PCI_BUS_ADDR_T_64BIT is set, in pci_bus_alloc_resource()
64bit address spaces (IORESOURCE_MEM_64) will be enabled which weren't enabled before.

This trivial/temporary hack fixes the problem:

--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -200,7 +200,7 @@ int pci_bus_alloc_resource(struct pci_bus *bus, struct resource *res,
                                           resource_size_t),
                 void *alignf_data)
  {
-#ifdef CONFIG_PCI_BUS_ADDR_T_64BIT
+#if defined(CONFIG_PCI_BUS_ADDR_T_64BIT) && !defined(CONFIG_PARISC)
         int rc;
  
         if (res->flags & IORESOURCE_MEM_64) {


but the real problem is probably, that the sym53c8xx driver or maybe the parisc PCI core code isn't 64bit clean?

I did some more debugging, and on parisc the first hang happens in function sym_check_raid() [called from sym2_probe()]
indrivers/scsi/sym53c8xx_2/sym_glue.c while trying to call readl():
   ram_val = readl(device->s.ramaddr + ram_size - 16);

Helge

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Yinghai Lu]

On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
> On 18.08.2015 20:44, Meelis Roos wrote:
>>> I did a bisect and commit 3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d ("PCI:
>>> Add pci_bus_addr_t") seems to be the culprit:
>>>
>>> http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d
>>
>> CC: patch author and linux-pci.
>
>
> I think this is the problem:
>
> --- a/drivers/pci/Kconfig
> +++ b/drivers/pci/Kconfig
> @@ -1,6 +1,10 @@
>  #
>  # PCI configuration
>  #
> +config PCI_BUS_ADDR_T_64BIT
> +       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
> +       depends on PCI

Thanks for bisecting.

Then we should change to

config PCI_BUS_ADDR_T_64BIT
       def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
       depends on PCI

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Meelis Roos]

> On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
> > On 18.08.2015 20:44, Meelis Roos wrote:
> >>> I did a bisect and commit 3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d ("PCI:
> >>> Add pci_bus_addr_t") seems to be the culprit:
> >>>
> >>> http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d
> >>
> >> CC: patch author and linux-pci.
> >
> >
> > I think this is the problem:
> >
> > --- a/drivers/pci/Kconfig
> > +++ b/drivers/pci/Kconfig
> > @@ -1,6 +1,10 @@
> >  #
> >  # PCI configuration
> >  #
> > +config PCI_BUS_ADDR_T_64BIT
> > +       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
> > +       depends on PCI
> 
> Thanks for bisecting.
> 
> Then we should change to
> 
> config PCI_BUS_ADDR_T_64BIT
>        def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
>        depends on PCI

Why SPARC64? The problem happened on parisc.

-- 
Meelis Roos (mroos@linux.ee)

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Yinghai Lu]

On Tue, Aug 18, 2015 at 9:48 PM, Meelis Roos <mroos@linux.ee> wrote:
>> On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
>>
>> Then we should change to
>>
>> config PCI_BUS_ADDR_T_64BIT
>>        def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
>>        depends on PCI
>
> Why SPARC64? The problem happened on parisc.
>

so will not set PCI_BUS_ADDR_T_64BIT for parisc.

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[John David Anglin]

On 2015-08-19, at 1:30 AM, Yinghai Lu wrote:

> On Tue, Aug 18, 2015 at 9:48 PM, Meelis Roos <mroos@linux.ee> wrote:
>>> On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
>>> 
>>> Then we should change to
>>> 
>>> config PCI_BUS_ADDR_T_64BIT
>>>       def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
>>>       depends on PCI
>> 
>> Why SPARC64? The problem happened on parisc.
>> 
> 
> so will not set PCI_BUS_ADDR_T_64BIT for parisc.

I'm not sure this is optimal.  While the A500 may only have 32-bit PCI, c8000 appears to have a mix
of 32 and 64-bit, and rp34XX is all 64-bit.

Dave
--
John David Anglin	dave.anglin@bell.net

Aw: Re: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Helge Deller]

> On 2015-08-19, at 1:30 AM, Yinghai Lu wrote:
> > On Tue, Aug 18, 2015 at 9:48 PM, Meelis Roos <mroos@linux.ee> wrote:
> >>> On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
> >>> 
> >>> Then we should change to
> >>> 
> >>> config PCI_BUS_ADDR_T_64BIT
> >>>       def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
> >>>       depends on PCI
> >> 
> >> Why SPARC64? The problem happened on parisc.
> > so will not set PCI_BUS_ADDR_T_64BIT for parisc.

I think given the current time frame it's probably the best approach to fix this problem for kernel 4.2.
It reverts the behaviour back to how it was before (for all arches beside SPARC64).
I'm still wondering if/why parisc is the only arch (in the sym53c8xx driver only!) which broke by this change...

> I'm not sure this is optimal.  While the A500 may only have 32-bit PCI, c8000 appears to have a mix
> of 32 and 64-bit, and rp34XX is all 64-bit.

True, but probably nobody of us noticed that we only used the 32-bit PCI interface even with 64bit kernel on parisc up to now?
pci_bus_alloc_resource() in drivers/pci/bus.c just disabled (flag=0) all 64bit resources for us.
But I agree, I think we need to fix drivers/parisc/lba_pci.c to correctly cope with 64bit pci addresses.

In the meantime I did some more debugging on sym_iomap_device() in drivers/scsi/sym53c8xx_2/sym_glue.c:
   pcibios_resource_to_bus(pdev->bus, &bus_addr,
                          &pdev->resource[i]);
   ram_base = bus_addr.start;

with (working) 32bit PCI addresses I get:
ressource = [mem 0xffffffff80002000-0xffffffff80003fff 64bit]
and pcibios_resource_to_bus() returning: ram_base = 0x80002000 

and with (failing) 64bit PCI addresses I get:
ressource = [mem 0xfffffff004000000-0xfffffff004001fff 64bit]
and pcibios_resource_to_bus() returning: ram_base = 0xfffffff004000000 

It seems the resource window doesn't get initialized correctly for 64bit...

Helge

Re: Re: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Bjorn Helgaas]

On Wed, Aug 19, 2015 at 04:06:38PM +0200, Helge Deller wrote:
> > On 2015-08-19, at 1:30 AM, Yinghai Lu wrote:
> > > On Tue, Aug 18, 2015 at 9:48 PM, Meelis Roos <mroos@linux.ee> wrote:
> > >>> On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
> > >>> 
> > >>> Then we should change to
> > >>> 
> > >>> config PCI_BUS_ADDR_T_64BIT
> > >>>       def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
> > >>>       depends on PCI
> > >> 
> > >> Why SPARC64? The problem happened on parisc.
> > > so will not set PCI_BUS_ADDR_T_64BIT for parisc.
> 
> I think given the current time frame it's probably the best approach to fix this problem for kernel 4.2.
> It reverts the behaviour back to how it was before (for all arches beside SPARC64).
> I'm still wondering if/why parisc is the only arch (in the sym53c8xx driver only!) which broke by this change...
> 
> > I'm not sure this is optimal.  While the A500 may only have 32-bit PCI, c8000 appears to have a mix
> > of 32 and 64-bit, and rp34XX is all 64-bit.
> 
> True, but probably nobody of us noticed that we only used the 32-bit PCI interface even with 64bit kernel on parisc up to now?
> pci_bus_alloc_resource() in drivers/pci/bus.c just disabled (flag=0) all 64bit resources for us.
> But I agree, I think we need to fix drivers/parisc/lba_pci.c to correctly cope with 64bit pci addresses.

I doubt we can fix the underlying issue before v4.2, but I'd at least
like to avoid it.  I applied the patch below to for-linus.

Yinghai proposed this:

> > >>>       def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)

I did this instead:

+	def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))

because (a) Yinghai's proposal suggests this change is related to
SPARC64, which is misleading, and (b) I want to keep 64-bit bus
addresses for all 64-bit kernels *except* PA-RISC.

I didn't mark it as Tested-by Meelis because I think he tested
Yinghai's proposal.  But I can add that back, given confirmation.


commit 51a660a2732c
Author: Bjorn Helgaas <bhelgaas@google.com>
Date:   Thu Aug 20 00:08:15 2015 -0500

    PCI: Don't use 64-bit bus addresses on PA-RISC
    
    Meelis and Helge reported that 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
    caused HPMCs on A500 and hangs on rp5470.
    
    PA-RISC does not set ARCH_DMA_ADDR_T_64BIT, even for 64-bit kernels, so
    prior to 3a9ad0b4fdcd, we always used 32-bit PCI addresses.  After
    3a9ad0b4fdcd, we do use 64-bit PCI addresses in 64-bit kernels, and
    apparently there's some PA-RISC problem related to them.
    
    Fixes: 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
    Link: http://lkml.kernel.org/r/alpine.LRH.2.11.1507260929000.30065@math.ut.ee
    Reported-by: Meelis Roos <mroos@linux.ee>
    Reported-by: Helge Deller <deller@gmx.de>
    Based-on-idea-by: Yinghai Lu <yinghai@kernel.org>
    Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
index 73de4ef..944f500 100644
--- a/drivers/pci/Kconfig
+++ b/drivers/pci/Kconfig
@@ -2,7 +2,7 @@
 # PCI configuration
 #
 config PCI_BUS_ADDR_T_64BIT
-	def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
+	def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))
 	depends on PCI
 
 config PCI_MSI

Re: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Helge Deller]

On 20.08.2015 07:42, Bjorn Helgaas wrote:
> On Wed, Aug 19, 2015 at 04:06:38PM +0200, Helge Deller wrote:
>>> On 2015-08-19, at 1:30 AM, Yinghai Lu wrote:
>>>> On Tue, Aug 18, 2015 at 9:48 PM, Meelis Roos <mroos@linux.ee> wrote:
>>>>>> On Tue, Aug 18, 2015 at 12:47 PM, Helge Deller <deller@gmx.de> wrote:
>>>>>>
>>>>>> Then we should change to
>>>>>>
>>>>>> config PCI_BUS_ADDR_T_64BIT
>>>>>>        def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
>>>>>>        depends on PCI
>>>>>
>>>>> Why SPARC64? The problem happened on parisc.
>>>> so will not set PCI_BUS_ADDR_T_64BIT for parisc.
>>
>> I think given the current time frame it's probably the best approach to fix this problem for kernel 4.2.
>> It reverts the behaviour back to how it was before (for all arches beside SPARC64).
>> I'm still wondering if/why parisc is the only arch (in the sym53c8xx driver only!) which broke by this change...
>>
>>> I'm not sure this is optimal.  While the A500 may only have 32-bit PCI, c8000 appears to have a mix
>>> of 32 and 64-bit, and rp34XX is all 64-bit.
>>
>> True, but probably nobody of us noticed that we only used the 32-bit PCI interface even with 64bit kernel on parisc up to now?
>> pci_bus_alloc_resource() in drivers/pci/bus.c just disabled (flag=0) all 64bit resources for us.
>> But I agree, I think we need to fix drivers/parisc/lba_pci.c to correctly cope with 64bit pci addresses.
>
> I doubt we can fix the underlying issue before v4.2,

Agreed.

> but I'd at least like to avoid it.  I applied the patch below to
> for-linus.
>
> Yinghai proposed this:
>
>>>>>>        def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
>
> I did this instead:
>
> +	def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))
>
> because (a) Yinghai's proposal suggests this change is related to
> SPARC64, which is misleading, and (b) I want to keep 64-bit bus
> addresses for all 64-bit kernels *except* PA-RISC.

Ok.
  
> I didn't mark it as Tested-by Meelis because I think he tested
> Yinghai's proposal.  But I can add that back, given confirmation.
>
>
> commit 51a660a2732c
> Author: Bjorn Helgaas <bhelgaas@google.com>
> Date:   Thu Aug 20 00:08:15 2015 -0500
>
>      PCI: Don't use 64-bit bus addresses on PA-RISC
>
>      Meelis and Helge reported that 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>      caused HPMCs on A500 and hangs on rp5470.
>
>      PA-RISC does not set ARCH_DMA_ADDR_T_64BIT, even for 64-bit kernels, so
>      prior to 3a9ad0b4fdcd, we always used 32-bit PCI addresses.  After
>      3a9ad0b4fdcd, we do use 64-bit PCI addresses in 64-bit kernels, and
>      apparently there's some PA-RISC problem related to them.
>
>      Fixes: 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>      Link: http://lkml.kernel.org/r/alpine.LRH.2.11.1507260929000.30065@math.ut.ee
>      Reported-by: Meelis Roos <mroos@linux.ee>
>      Reported-by: Helge Deller <deller@gmx.de>
>      Based-on-idea-by: Yinghai Lu <yinghai@kernel.org>
>      Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

Looks OK. I will test later today.

Can you please add
  CC: stable@vger.kernel.org  # v3.19+
since commit 3a9ad0b4fdcd had that too and now all kernels down to 3.19+ are
broken on parisc...

Thanks!
Helge  
  
> diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
> index 73de4ef..944f500 100644
> --- a/drivers/pci/Kconfig
> +++ b/drivers/pci/Kconfig
> @@ -2,7 +2,7 @@
>   # PCI configuration
>   #
>   config PCI_BUS_ADDR_T_64BIT
> -	def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
> +	def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))
>   	depends on PCI
>
>   config PCI_MSI

Re: Re: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Yinghai Lu]

On Wed, Aug 19, 2015 at 10:42 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:

> commit 51a660a2732c
> Author: Bjorn Helgaas <bhelgaas@google.com>
> Date:   Thu Aug 20 00:08:15 2015 -0500
>
>     PCI: Don't use 64-bit bus addresses on PA-RISC
>
>     Meelis and Helge reported that 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>     caused HPMCs on A500 and hangs on rp5470.
>
>     PA-RISC does not set ARCH_DMA_ADDR_T_64BIT, even for 64-bit kernels, so
>     prior to 3a9ad0b4fdcd, we always used 32-bit PCI addresses.  After
>     3a9ad0b4fdcd, we do use 64-bit PCI addresses in 64-bit kernels, and
>     apparently there's some PA-RISC problem related to them.
>
>     Fixes: 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>     Link: http://lkml.kernel.org/r/alpine.LRH.2.11.1507260929000.30065@math.ut.ee
>     Reported-by: Meelis Roos <mroos@linux.ee>
>     Reported-by: Helge Deller <deller@gmx.de>
>     Based-on-idea-by: Yinghai Lu <yinghai@kernel.org>
>     Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
>
> diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
> index 73de4ef..944f500 100644
> --- a/drivers/pci/Kconfig
> +++ b/drivers/pci/Kconfig
> @@ -2,7 +2,7 @@
>  # PCI configuration
>  #
>  config PCI_BUS_ADDR_T_64BIT
> -       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
> +       def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))
>         depends on PCI
>
>  config PCI_MSI

Yes, that is better.

Acked-by: Yinghai Lu <yinghai@kernel.org>

Re: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Helge Deller]

On 20.08.2015 21:31, Yinghai Lu wrote:
> On Wed, Aug 19, 2015 at 10:42 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>
>> commit 51a660a2732c
>> Author: Bjorn Helgaas <bhelgaas@google.com>
>> Date:   Thu Aug 20 00:08:15 2015 -0500
>>
>>      PCI: Don't use 64-bit bus addresses on PA-RISC
>>
>>      Meelis and Helge reported that 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>>      caused HPMCs on A500 and hangs on rp5470.
>>
>>      PA-RISC does not set ARCH_DMA_ADDR_T_64BIT, even for 64-bit kernels, so
>>      prior to 3a9ad0b4fdcd, we always used 32-bit PCI addresses.  After
>>      3a9ad0b4fdcd, we do use 64-bit PCI addresses in 64-bit kernels, and
>>      apparently there's some PA-RISC problem related to them.
>>
>>      Fixes: 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>>      Link: http://lkml.kernel.org/r/alpine.LRH.2.11.1507260929000.30065@math.ut.ee
>>      Reported-by: Meelis Roos <mroos@linux.ee>
>>      Reported-by: Helge Deller <deller@gmx.de>
>>      Based-on-idea-by: Yinghai Lu <yinghai@kernel.org>
>>      Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
>>
>> diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
>> index 73de4ef..944f500 100644
>> --- a/drivers/pci/Kconfig
>> +++ b/drivers/pci/Kconfig
>> @@ -2,7 +2,7 @@
>>   # PCI configuration
>>   #
>>   config PCI_BUS_ADDR_T_64BIT
>> -       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
>> +       def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))
>>          depends on PCI
>>
>>   config PCI_MSI
>
> Yes, that is better.
>
> Acked-by: Yinghai Lu <yinghai@kernel.org>


Tested-by: Helge Deller <deller@gmx.de>
   and please add:
CC: stable@vger.kernel.org  # v3.19+

Helge

Re: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Bjorn Helgaas]

On Thu, Aug 20, 2015 at 2:15 PM, Helge Deller <deller@gmx.de> wrote:
> On 20.08.2015 21:31, Yinghai Lu wrote:
>>
>> On Wed, Aug 19, 2015 at 10:42 PM, Bjorn Helgaas <bhelgaas@google.com>
>> wrote:
>>
>>> commit 51a660a2732c
>>> Author: Bjorn Helgaas <bhelgaas@google.com>
>>> Date:   Thu Aug 20 00:08:15 2015 -0500
>>>
>>>      PCI: Don't use 64-bit bus addresses on PA-RISC
>>>
>>>      Meelis and Helge reported that 3a9ad0b4fdcd ("PCI: Add
>>> pci_bus_addr_t")
>>>      caused HPMCs on A500 and hangs on rp5470.
>>>
>>>      PA-RISC does not set ARCH_DMA_ADDR_T_64BIT, even for 64-bit kernels,
>>> so
>>>      prior to 3a9ad0b4fdcd, we always used 32-bit PCI addresses.  After
>>>      3a9ad0b4fdcd, we do use 64-bit PCI addresses in 64-bit kernels, and
>>>      apparently there's some PA-RISC problem related to them.
>>>
>>>      Fixes: 3a9ad0b4fdcd ("PCI: Add pci_bus_addr_t")
>>>      Link:
>>> http://lkml.kernel.org/r/alpine.LRH.2.11.1507260929000.30065@math.ut.ee
>>>      Reported-by: Meelis Roos <mroos@linux.ee>
>>>      Reported-by: Helge Deller <deller@gmx.de>
>>>      Based-on-idea-by: Yinghai Lu <yinghai@kernel.org>
>>>      Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
>>>
>>> diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
>>> index 73de4ef..944f500 100644
>>> --- a/drivers/pci/Kconfig
>>> +++ b/drivers/pci/Kconfig
>>> @@ -2,7 +2,7 @@
>>>   # PCI configuration
>>>   #
>>>   config PCI_BUS_ADDR_T_64BIT
>>> -       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
>>> +       def_bool y if (ARCH_DMA_ADDR_T_64BIT || (64BIT && !PARISC))
>>>          depends on PCI
>>>
>>>   config PCI_MSI
>>
>>
>> Yes, that is better.
>>
>> Acked-by: Yinghai Lu <yinghai@kernel.org>
>
>
>
> Tested-by: Helge Deller <deller@gmx.de>
>   and please add:
> CC: stable@vger.kernel.org  # v3.19+

Done, thanks for testing this!

Re: Aw: A500 boot crash in 4.2.0-rc3-00246-g763e326

[Meelis Roos]

> >>> I did a bisect and commit 3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d ("PCI:
> >>> Add pci_bus_addr_t") seems to be the culprit:
> >>>
> >>> http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=3a9ad0b4fdcd57f775d3615004c8c64c021a9e7d
> >>
> >> CC: patch author and linux-pci.
> >
> >
> > I think this is the problem:
> >
> > --- a/drivers/pci/Kconfig
> > +++ b/drivers/pci/Kconfig
> > @@ -1,6 +1,10 @@
> >  #
> >  # PCI configuration
> >  #
> > +config PCI_BUS_ADDR_T_64BIT
> > +       def_bool y if (ARCH_DMA_ADDR_T_64BIT || 64BIT)
> > +       depends on PCI
> 
> Thanks for bisecting.
> 
> Then we should change to
> 
> config PCI_BUS_ADDR_T_64BIT
>        def_bool y if (ARCH_DMA_ADDR_T_64BIT || SPARC64)
>        depends on PCI

This works for my A500.

-- 
Meelis Roos (mroos@linux.ee)