linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup
@ 2016-12-22  9:07 Zhou Wang
  2017-01-03  1:28 ` Zhou Wang
                   ` (2 more replies)
  0 siblings, 3 replies; 6+ messages in thread
From: Zhou Wang @ 2016-12-22  9:07 UTC (permalink / raw)
  To: Rafael J. Wysocki, Len Brown, Tomasz Nowicki, Jayachandran C,
	Lorenzo Pieralisi, jorn Helgaas
  Cc: liudongdong3, gabriele.paoloni, linux-acpi, linux-pci,
	linux-kernel, Zhou Wang

Multiple PCIe host bridges may exists in one PCIe segment. So bus range for each
host bridge should be in the coverage of bus range of related PCIe segment.

This patch will support this kind of scenario:

MCFG:
	bus range: 0x00~0xff.
	segment: 0.
DSDT:
	host bridge 1:
		bus range: 0x00~0x1f.
		segment: 0.
	host bridge 2:
		bus range: 0x20~0x4f.
		segment: 0.

Signed-off-by: Zhou Wang <wangzhou1@hisilicon.com>
---
 drivers/acpi/pci_mcfg.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c
index b5b376e..46a3e32 100644
--- a/drivers/acpi/pci_mcfg.c
+++ b/drivers/acpi/pci_mcfg.c
@@ -40,11 +40,10 @@ phys_addr_t pci_mcfg_lookup(u16 seg, struct resource *bus_res)
 	struct mcfg_entry *e;
 
 	/*
-	 * We expect exact match, unless MCFG entry end bus covers more than
-	 * specified by caller.
+	 * We expect the range in bus_res in the coverage of MCFG bus range.
 	 */
 	list_for_each_entry(e, &pci_mcfg_list, list) {
-		if (e->segment == seg && e->bus_start == bus_res->start &&
+		if (e->segment == seg && e->bus_start <= bus_res->start &&
 		    e->bus_end >= bus_res->end)
 			return e->addr;
 	}
-- 
1.9.1

^ permalink raw reply related	[flat|nested] 6+ messages in thread
* Re: [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup
  2016-12-22  9:07 [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup Zhou Wang
@ 2017-01-03  1:28 ` Zhou Wang
  2017-01-03  6:39 ` Tomasz Nowicki
  2017-01-03 12:00 ` Lorenzo Pieralisi
  2 siblings, 0 replies; 6+ messages in thread
From: Zhou Wang @ 2017-01-03  1:28 UTC (permalink / raw)
  To: Rafael J. Wysocki, Len Brown, Tomasz Nowicki, Jayachandran C,
	Lorenzo Pieralisi, jorn Helgaas
  Cc: liudongdong3, gabriele.paoloni, linux-acpi, linux-pci, linux-kernel

On 2016/12/22 17:07, Zhou Wang wrote:
> Multiple PCIe host bridges may exists in one PCIe segment. So bus range for each
> host bridge should be in the coverage of bus range of related PCIe segment.
> 
> This patch will support this kind of scenario:
> 
> MCFG:
> 	bus range: 0x00~0xff.
> 	segment: 0.
> DSDT:
> 	host bridge 1:
> 		bus range: 0x00~0x1f.
> 		segment: 0.
> 	host bridge 2:
> 		bus range: 0x20~0x4f.
> 		segment: 0.
> 
> Signed-off-by: Zhou Wang <wangzhou1@hisilicon.com>
> ---
>  drivers/acpi/pci_mcfg.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c
> index b5b376e..46a3e32 100644
> --- a/drivers/acpi/pci_mcfg.c
> +++ b/drivers/acpi/pci_mcfg.c
> @@ -40,11 +40,10 @@ phys_addr_t pci_mcfg_lookup(u16 seg, struct resource *bus_res)
>  	struct mcfg_entry *e;
>  
>  	/*
> -	 * We expect exact match, unless MCFG entry end bus covers more than
> -	 * specified by caller.
> +	 * We expect the range in bus_res in the coverage of MCFG bus range.
>  	 */
>  	list_for_each_entry(e, &pci_mcfg_list, list) {
> -		if (e->segment == seg && e->bus_start == bus_res->start &&
> +		if (e->segment == seg && e->bus_start <= bus_res->start &&
>  		    e->bus_end >= bus_res->end)
>  			return e->addr;
>  	}
> 

Any idea about this RFC patch?

Thanks,
Zhou

.

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup
  2016-12-22  9:07 [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup Zhou Wang
  2017-01-03  1:28 ` Zhou Wang
@ 2017-01-03  6:39 ` Tomasz Nowicki
  2017-01-03  9:07   ` Zhou Wang
  2017-01-03 12:00 ` Lorenzo Pieralisi
  2 siblings, 1 reply; 6+ messages in thread
From: Tomasz Nowicki @ 2017-01-03  6:39 UTC (permalink / raw)
  To: Zhou Wang, Rafael J. Wysocki, Len Brown, Jayachandran C,
	Lorenzo Pieralisi, jorn Helgaas
  Cc: liudongdong3, gabriele.paoloni, linux-acpi, linux-pci, linux-kernel

On 22.12.2016 10:07, Zhou Wang wrote:
> Multiple PCIe host bridges may exists in one PCIe segment. So bus range for each
> host bridge should be in the coverage of bus range of related PCIe segment.
>
> This patch will support this kind of scenario:
>
> MCFG:
> 	bus range: 0x00~0xff.
> 	segment: 0.
> DSDT:
> 	host bridge 1:
> 		bus range: 0x00~0x1f.
> 		segment: 0.
> 	host bridge 2:
> 		bus range: 0x20~0x4f.
> 		segment: 0.
>
> Signed-off-by: Zhou Wang <wangzhou1@hisilicon.com>
> ---
>  drivers/acpi/pci_mcfg.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c
> index b5b376e..46a3e32 100644
> --- a/drivers/acpi/pci_mcfg.c
> +++ b/drivers/acpi/pci_mcfg.c
> @@ -40,11 +40,10 @@ phys_addr_t pci_mcfg_lookup(u16 seg, struct resource *bus_res)
>  	struct mcfg_entry *e;
>
>  	/*
> -	 * We expect exact match, unless MCFG entry end bus covers more than
> -	 * specified by caller.
> +	 * We expect the range in bus_res in the coverage of MCFG bus range.
>  	 */
>  	list_for_each_entry(e, &pci_mcfg_list, list) {
> -		if (e->segment == seg && e->bus_start == bus_res->start &&
> +		if (e->segment == seg && e->bus_start <= bus_res->start &&
>  		    e->bus_end >= bus_res->end)
>  			return e->addr;
>  	}
>

Looks good to me.

Reviewed-by: Tomasz Nowicki <tn@semihalf.com>

Thanks,
Tomasz

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup
  2017-01-03  6:39 ` Tomasz Nowicki
@ 2017-01-03  9:07   ` Zhou Wang
  0 siblings, 0 replies; 6+ messages in thread
From: Zhou Wang @ 2017-01-03  9:07 UTC (permalink / raw)
  To: Tomasz Nowicki, Rafael J. Wysocki, Len Brown, Jayachandran C,
	Lorenzo Pieralisi, jorn Helgaas
  Cc: liudongdong3, gabriele.paoloni, linux-acpi, linux-pci, linux-kernel

On 2017/1/3 14:39, Tomasz Nowicki wrote:
> On 22.12.2016 10:07, Zhou Wang wrote:
>> Multiple PCIe host bridges may exists in one PCIe segment. So bus range for each
>> host bridge should be in the coverage of bus range of related PCIe segment.
>>
>> This patch will support this kind of scenario:
>>
>> MCFG:
>>     bus range: 0x00~0xff.
>>     segment: 0.
>> DSDT:
>>     host bridge 1:
>>         bus range: 0x00~0x1f.
>>         segment: 0.
>>     host bridge 2:
>>         bus range: 0x20~0x4f.
>>         segment: 0.
>>
>> Signed-off-by: Zhou Wang <wangzhou1@hisilicon.com>
>> ---
>>  drivers/acpi/pci_mcfg.c | 5 ++---
>>  1 file changed, 2 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c
>> index b5b376e..46a3e32 100644
>> --- a/drivers/acpi/pci_mcfg.c
>> +++ b/drivers/acpi/pci_mcfg.c
>> @@ -40,11 +40,10 @@ phys_addr_t pci_mcfg_lookup(u16 seg, struct resource *bus_res)
>>      struct mcfg_entry *e;
>>
>>      /*
>> -     * We expect exact match, unless MCFG entry end bus covers more than
>> -     * specified by caller.
>> +     * We expect the range in bus_res in the coverage of MCFG bus range.
>>       */
>>      list_for_each_entry(e, &pci_mcfg_list, list) {
>> -        if (e->segment == seg && e->bus_start == bus_res->start &&
>> +        if (e->segment == seg && e->bus_start <= bus_res->start &&
>>              e->bus_end >= bus_res->end)
>>              return e->addr;
>>      }
>>
> 
> Looks good to me.
> 
> Reviewed-by: Tomasz Nowicki <tn@semihalf.com>

Hi Tomasz,

Thanks for your review.

Regards,
Zhou

> 
> Thanks,
> Tomasz
> 
> .
> 

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup
  2016-12-22  9:07 [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup Zhou Wang
  2017-01-03  1:28 ` Zhou Wang
  2017-01-03  6:39 ` Tomasz Nowicki
@ 2017-01-03 12:00 ` Lorenzo Pieralisi
  2017-01-04  5:44   ` Zhou Wang
  2 siblings, 1 reply; 6+ messages in thread
From: Lorenzo Pieralisi @ 2017-01-03 12:00 UTC (permalink / raw)
  To: Zhou Wang
  Cc: Rafael J. Wysocki, Len Brown, Tomasz Nowicki, Jayachandran C,
	jorn Helgaas, liudongdong3, gabriele.paoloni, linux-acpi,
	linux-pci, linux-kernel

On Thu, Dec 22, 2016 at 05:07:43PM +0800, Zhou Wang wrote:
> Multiple PCIe host bridges may exists in one PCIe segment. So bus range for each
> host bridge should be in the coverage of bus range of related PCIe segment.
> 
> This patch will support this kind of scenario:
> 
> MCFG:
> 	bus range: 0x00~0xff.
> 	segment: 0.
> DSDT:
> 	host bridge 1:
> 		bus range: 0x00~0x1f.
> 		segment: 0.
> 	host bridge 2:
> 		bus range: 0x20~0x4f.
> 		segment: 0.

"The configuration data provided by an MCFG region (ie PCI segment and
bus range) may span multiple host bridges.

Current code in pci_mcfg_lookup() carries out an exact match of host
bridge bus range start value against the MCFG region(s) bus range start
value which would cause configurations like the following:

MCFG region:
	bus range: 0x00~0xff.
	segment: 0.

PCI host bridges configuration (segment numbers and bus ranges):
	host bridge 1:
		bus range: 0x00~0x1f.
		segment: 0.
	host bridge 2:
		bus range: 0x20~0x4f.
		segment: 0.

to fail, in that the bus range start value for host bridge 2 does
not match the bus range start value of the respective MCFG region.

Relax the bus range check in pci_mcfg_lookup() to cater for
PCI configurations with multiple host bridges sharing the same
MCFG region."

Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>

> Signed-off-by: Zhou Wang <wangzhou1@hisilicon.com>
> ---
>  drivers/acpi/pci_mcfg.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c
> index b5b376e..46a3e32 100644
> --- a/drivers/acpi/pci_mcfg.c
> +++ b/drivers/acpi/pci_mcfg.c
> @@ -40,11 +40,10 @@ phys_addr_t pci_mcfg_lookup(u16 seg, struct resource *bus_res)
>  	struct mcfg_entry *e;
>  
>  	/*
> -	 * We expect exact match, unless MCFG entry end bus covers more than
> -	 * specified by caller.
> +	 * We expect the range in bus_res in the coverage of MCFG bus range.
>  	 */
>  	list_for_each_entry(e, &pci_mcfg_list, list) {
> -		if (e->segment == seg && e->bus_start == bus_res->start &&
> +		if (e->segment == seg && e->bus_start <= bus_res->start &&
>  		    e->bus_end >= bus_res->end)
>  			return e->addr;
>  	}
> -- 
> 1.9.1
> 

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup
  2017-01-03 12:00 ` Lorenzo Pieralisi
@ 2017-01-04  5:44   ` Zhou Wang
  0 siblings, 0 replies; 6+ messages in thread
From: Zhou Wang @ 2017-01-04  5:44 UTC (permalink / raw)
  To: Lorenzo Pieralisi
  Cc: Rafael J. Wysocki, Len Brown, Tomasz Nowicki, Jayachandran C,
	jorn Helgaas, liudongdong3, gabriele.paoloni, linux-acpi,
	linux-pci, linux-kernel

On 2017/1/3 20:00, Lorenzo Pieralisi wrote:
> On Thu, Dec 22, 2016 at 05:07:43PM +0800, Zhou Wang wrote:
>> Multiple PCIe host bridges may exists in one PCIe segment. So bus range for each
>> host bridge should be in the coverage of bus range of related PCIe segment.
>>
>> This patch will support this kind of scenario:
>>
>> MCFG:
>> 	bus range: 0x00~0xff.
>> 	segment: 0.
>> DSDT:
>> 	host bridge 1:
>> 		bus range: 0x00~0x1f.
>> 		segment: 0.
>> 	host bridge 2:
>> 		bus range: 0x20~0x4f.
>> 		segment: 0.
> 
> "The configuration data provided by an MCFG region (ie PCI segment and
> bus range) may span multiple host bridges.
> 
> Current code in pci_mcfg_lookup() carries out an exact match of host
> bridge bus range start value against the MCFG region(s) bus range start
> value which would cause configurations like the following:
> 
> MCFG region:
> 	bus range: 0x00~0xff.
> 	segment: 0.
> 
> PCI host bridges configuration (segment numbers and bus ranges):
> 	host bridge 1:
> 		bus range: 0x00~0x1f.
> 		segment: 0.
> 	host bridge 2:
> 		bus range: 0x20~0x4f.
> 		segment: 0.
> 
> to fail, in that the bus range start value for host bridge 2 does
> not match the bus range start value of the respective MCFG region.
> 
> Relax the bus range check in pci_mcfg_lookup() to cater for
> PCI configurations with multiple host bridges sharing the same
> MCFG region."
> 
> Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>

Hi Lorenzo,

Thanks for your view.

I will modify the commit message and post a new version patch.

Regards,
Zhou

> 
>> Signed-off-by: Zhou Wang <wangzhou1@hisilicon.com>
>> ---
>>  drivers/acpi/pci_mcfg.c | 5 ++---
>>  1 file changed, 2 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c
>> index b5b376e..46a3e32 100644
>> --- a/drivers/acpi/pci_mcfg.c
>> +++ b/drivers/acpi/pci_mcfg.c
>> @@ -40,11 +40,10 @@ phys_addr_t pci_mcfg_lookup(u16 seg, struct resource *bus_res)
>>  	struct mcfg_entry *e;
>>  
>>  	/*
>> -	 * We expect exact match, unless MCFG entry end bus covers more than
>> -	 * specified by caller.
>> +	 * We expect the range in bus_res in the coverage of MCFG bus range.
>>  	 */
>>  	list_for_each_entry(e, &pci_mcfg_list, list) {
>> -		if (e->segment == seg && e->bus_start == bus_res->start &&
>> +		if (e->segment == seg && e->bus_start <= bus_res->start &&
>>  		    e->bus_end >= bus_res->end)
>>  			return e->addr;
>>  	}
>> -- 
>> 1.9.1
>>
> 
> .
> 

^ permalink raw reply	[flat|nested] 6+ messages in thread
end of thread, other threads:[~2017-01-04  5:47 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-12-22  9:07 [RFC PATCH] ACPI/PCI: Fix bus range comparation in pci_mcfg_lookup Zhou Wang
2017-01-03  1:28 ` Zhou Wang
2017-01-03  6:39 ` Tomasz Nowicki
2017-01-03  9:07   ` Zhou Wang
2017-01-03 12:00 ` Lorenzo Pieralisi
2017-01-04  5:44   ` Zhou Wang

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).