xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] xen/pciback: Fix conf_space read/write overlap check.
@ 2016-06-21 14:37 Andrey Grodzovsky
  2016-06-21 15:52 ` Boris Ostrovsky
                   ` (2 more replies)
  0 siblings, 3 replies; 8+ messages in thread
From: Andrey Grodzovsky @ 2016-06-21 14:37 UTC (permalink / raw)
  To: xen-devel; +Cc: Boris Ostrovsky, stable, Jan Beulich, Andrey Grodzovsky, jw

Current overlap check is evaluating to false a case where a filter field
is fully contained (proper subset) of a r/w request.
This change applies classical overlap check instead to include
all the scenarios.

Related to https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html

Cc: Jan Beulich <JBeulich@suse.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: stable@vger.kernel.org
Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>
---
 drivers/xen/xen-pciback/conf_space.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c
index 8e67336..6a25533 100644
--- a/drivers/xen/xen-pciback/conf_space.c
+++ b/drivers/xen/xen-pciback/conf_space.c
@@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int offset, int size,
 		field_start = OFFSET(cfg_entry);
 		field_end = OFFSET(cfg_entry) + field->size;
 
-		if ((req_start >= field_start && req_start < field_end)
-		    || (req_end > field_start && req_end <= field_end)) {
+		 if (req_end > field_start && field_end > req_start) {
 			err = conf_space_read(dev, cfg_entry, field_start,
 					      &tmp_val);
 			if (err)
@@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value)
 		field_start = OFFSET(cfg_entry);
 		field_end = OFFSET(cfg_entry) + field->size;
 
-		if ((req_start >= field_start && req_start < field_end)
-		    || (req_end > field_start && req_end <= field_end)) {
+		 if (req_end > field_start && field_end > req_start) {
 			tmp_val = 0;
 
 			err = xen_pcibk_config_read(dev, field_start,
-- 
2.1.4


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply related	[flat|nested] 8+ messages in thread

* Re: [PATCH] xen/pciback: Fix conf_space read/write overlap check.
  2016-06-21 14:37 [PATCH] xen/pciback: Fix conf_space read/write overlap check Andrey Grodzovsky
@ 2016-06-21 15:52 ` Boris Ostrovsky
  2016-06-22  4:11   ` Juergen Gross
  2016-06-21 17:15 ` David Vrabel
       [not found] ` <57697614.5020003@citrix.com>
  2 siblings, 1 reply; 8+ messages in thread
From: Boris Ostrovsky @ 2016-06-21 15:52 UTC (permalink / raw)
  To: Andrey Grodzovsky, xen-devel
  Cc: Juergen Gross, jw, Linux Kernel Mailing List, stable,
	David Vrabel, Jan Beulich

On 06/21/2016 10:37 AM, Andrey Grodzovsky wrote:
> Current overlap check is evaluating to false a case where a filter field
> is fully contained (proper subset) of a r/w request.
> This change applies classical overlap check instead to include
> all the scenarios.
>
> Related to https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html
>
> Cc: Jan Beulich <JBeulich@suse.com>
> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>

+ David and Juergen (maintainers) and kernel list.

Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>


> ---
>  drivers/xen/xen-pciback/conf_space.c | 6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c
> index 8e67336..6a25533 100644
> --- a/drivers/xen/xen-pciback/conf_space.c
> +++ b/drivers/xen/xen-pciback/conf_space.c
> @@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int offset, int size,
>  		field_start = OFFSET(cfg_entry);
>  		field_end = OFFSET(cfg_entry) + field->size;
>  
> -		if ((req_start >= field_start && req_start < field_end)
> -		    || (req_end > field_start && req_end <= field_end)) {
> +		 if (req_end > field_start && field_end > req_start) {
>  			err = conf_space_read(dev, cfg_entry, field_start,
>  					      &tmp_val);
>  			if (err)
> @@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value)
>  		field_start = OFFSET(cfg_entry);
>  		field_end = OFFSET(cfg_entry) + field->size;
>  
> -		if ((req_start >= field_start && req_start < field_end)
> -		    || (req_end > field_start && req_end <= field_end)) {
> +		 if (req_end > field_start && field_end > req_start) {
>  			tmp_val = 0;
>  
>  			err = xen_pcibk_config_read(dev, field_start,



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] xen/pciback: Fix conf_space read/write overlap check.
  2016-06-21 14:37 [PATCH] xen/pciback: Fix conf_space read/write overlap check Andrey Grodzovsky
  2016-06-21 15:52 ` Boris Ostrovsky
@ 2016-06-21 17:15 ` David Vrabel
       [not found] ` <57697614.5020003@citrix.com>
  2 siblings, 0 replies; 8+ messages in thread
From: David Vrabel @ 2016-06-21 17:15 UTC (permalink / raw)
  To: Andrey Grodzovsky, xen-devel; +Cc: jw, Boris Ostrovsky, Jan Beulich, stable

On 21/06/16 15:37, Andrey Grodzovsky wrote:
> Current overlap check is evaluating to false a case where a filter field
> is fully contained (proper subset) of a r/w request.
> This change applies classical overlap check instead to include
> all the scenarios.

Reviewed-by: David Vrabel <david.vrabel@citrix.com>

But the commit message could do with a concrete example of an access
that failed.

David

> 
> Related to https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html
> 
> Cc: Jan Beulich <JBeulich@suse.com>
> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>
> ---
>  drivers/xen/xen-pciback/conf_space.c | 6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c
> index 8e67336..6a25533 100644
> --- a/drivers/xen/xen-pciback/conf_space.c
> +++ b/drivers/xen/xen-pciback/conf_space.c
> @@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int offset, int size,
>  		field_start = OFFSET(cfg_entry);
>  		field_end = OFFSET(cfg_entry) + field->size;
>  
> -		if ((req_start >= field_start && req_start < field_end)
> -		    || (req_end > field_start && req_end <= field_end)) {
> +		 if (req_end > field_start && field_end > req_start) {
>  			err = conf_space_read(dev, cfg_entry, field_start,
>  					      &tmp_val);
>  			if (err)
> @@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value)
>  		field_start = OFFSET(cfg_entry);
>  		field_end = OFFSET(cfg_entry) + field->size;
>  
> -		if ((req_start >= field_start && req_start < field_end)
> -		    || (req_end > field_start && req_end <= field_end)) {
> +		 if (req_end > field_start && field_end > req_start) {
>  			tmp_val = 0;
>  
>  			err = xen_pcibk_config_read(dev, field_start,
> 


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] xen/pciback: Fix conf_space read/write overlap check.
       [not found] ` <57697614.5020003@citrix.com>
@ 2016-06-21 18:20   ` Andrey Grodzovsky
  0 siblings, 0 replies; 8+ messages in thread
From: Andrey Grodzovsky @ 2016-06-21 18:20 UTC (permalink / raw)
  To: David Vrabel
  Cc: Jürgen Walter • Quattru, xen-devel, Boris Ostrovsky,
	Jan Beulich, stable


[-- Attachment #1.1: Type: text/plain, Size: 2366 bytes --]

On Tue, Jun 21, 2016 at 1:15 PM, David Vrabel <david.vrabel@citrix.com>
wrote:

> On 21/06/16 15:37, Andrey Grodzovsky wrote:
> > Current overlap check is evaluating to false a case where a filter field
> > is fully contained (proper subset) of a r/w request.
> > This change applies classical overlap check instead to include
> > all the scenarios.
>
> Reviewed-by: David Vrabel <david.vrabel@citrix.com>
>
> But the commit message could do with a concrete example of an access
> that failed.
>
> David


Thanks, will update the commit message and resubmit.

Andrey

>



>
> > Related to
> https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html
> >
> > Cc: Jan Beulich <JBeulich@suse.com>
> > Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>
> > ---
> >  drivers/xen/xen-pciback/conf_space.c | 6 ++----
> >  1 file changed, 2 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/xen/xen-pciback/conf_space.c
> b/drivers/xen/xen-pciback/conf_space.c
> > index 8e67336..6a25533 100644
> > --- a/drivers/xen/xen-pciback/conf_space.c
> > +++ b/drivers/xen/xen-pciback/conf_space.c
> > @@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int
> offset, int size,
> >               field_start = OFFSET(cfg_entry);
> >               field_end = OFFSET(cfg_entry) + field->size;
> >
> > -             if ((req_start >= field_start && req_start < field_end)
> > -                 || (req_end > field_start && req_end <= field_end)) {
> > +              if (req_end > field_start && field_end > req_start) {
> >                       err = conf_space_read(dev, cfg_entry, field_start,
> >                                             &tmp_val);
> >                       if (err)
> > @@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int
> offset, int size, u32 value)
> >               field_start = OFFSET(cfg_entry);
> >               field_end = OFFSET(cfg_entry) + field->size;
> >
> > -             if ((req_start >= field_start && req_start < field_end)
> > -                 || (req_end > field_start && req_end <= field_end)) {
> > +              if (req_end > field_start && field_end > req_start) {
> >                       tmp_val = 0;
> >
> >                       err = xen_pcibk_config_read(dev, field_start,
> >
>
>

[-- Attachment #1.2: Type: text/html, Size: 4053 bytes --]

[-- Attachment #2: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] xen/pciback: Fix conf_space read/write overlap check.
  2016-06-21 15:52 ` Boris Ostrovsky
@ 2016-06-22  4:11   ` Juergen Gross
  0 siblings, 0 replies; 8+ messages in thread
From: Juergen Gross @ 2016-06-22  4:11 UTC (permalink / raw)
  To: Boris Ostrovsky, Andrey Grodzovsky, xen-devel
  Cc: stable, Linux Kernel Mailing List, Jan Beulich, jw, David Vrabel

On 21/06/16 17:52, Boris Ostrovsky wrote:
> On 06/21/2016 10:37 AM, Andrey Grodzovsky wrote:
>> Current overlap check is evaluating to false a case where a filter field
>> is fully contained (proper subset) of a r/w request.
>> This change applies classical overlap check instead to include
>> all the scenarios.
>>
>> Related to https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html
>>
>> Cc: Jan Beulich <JBeulich@suse.com>
>> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
>> Cc: stable@vger.kernel.org
>> Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>
> 
> + David and Juergen (maintainers) and kernel list.
> 
> Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>

Acked-by: Juergen Gross <jgross@suse.com>

> 
> 
>> ---
>>  drivers/xen/xen-pciback/conf_space.c | 6 ++----
>>  1 file changed, 2 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c
>> index 8e67336..6a25533 100644
>> --- a/drivers/xen/xen-pciback/conf_space.c
>> +++ b/drivers/xen/xen-pciback/conf_space.c
>> @@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int offset, int size,
>>  		field_start = OFFSET(cfg_entry);
>>  		field_end = OFFSET(cfg_entry) + field->size;
>>  
>> -		if ((req_start >= field_start && req_start < field_end)
>> -		    || (req_end > field_start && req_end <= field_end)) {
>> +		 if (req_end > field_start && field_end > req_start) {
>>  			err = conf_space_read(dev, cfg_entry, field_start,
>>  					      &tmp_val);
>>  			if (err)
>> @@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value)
>>  		field_start = OFFSET(cfg_entry);
>>  		field_end = OFFSET(cfg_entry) + field->size;
>>  
>> -		if ((req_start >= field_start && req_start < field_end)
>> -		    || (req_end > field_start && req_end <= field_end)) {
>> +		 if (req_end > field_start && field_end > req_start) {
>>  			tmp_val = 0;
>>  
>>  			err = xen_pcibk_config_read(dev, field_start,
> 
> 
> 


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] xen/pciback: Fix conf_space read/write overlap check.
       [not found] <1466533596-15122-1-git-send-email-andrey2805@gmail.com>
  2016-06-22  6:49 ` Jan Beulich
@ 2016-06-23  9:56 ` David Vrabel
  1 sibling, 0 replies; 8+ messages in thread
From: David Vrabel @ 2016-06-23  9:56 UTC (permalink / raw)
  To: Andrey Grodzovsky, xen-devel
  Cc: Boris Ostrovsky, Jan Beulich, stable, jw, david.vrabel

On 21/06/16 19:26, Andrey Grodzovsky wrote:
> Current overlap check is evaluating to false a case where a filter field
> is fully contained (proper subset) of a r/w request.
> This change applies classical overlap check instead to include
> all the scenarios.
> 
> More specifically, for (Hilscher GmbH CIFX 50E-DP(M/S)) device
> driver the logic is such that the entire confspace  is read and
> written in 4 byte chunks.In this case as an example, CACHE_LINE_SIZE,
> LATENCY_TIMER and PCI_BIST are arriving together in one call to
> xen_pcibk_config_write with offset == 0xc and size == 4.
> With the exsisting overlap check LATENCY_TIMER field
> (offset == 0xd, length == 1) is fully contained in the write request
> and hence is excluded from write, which is incorrect.

Applied to for-linus-4.7b, thanks.

David

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH] xen/pciback: Fix conf_space read/write overlap check.
       [not found] <1466533596-15122-1-git-send-email-andrey2805@gmail.com>
@ 2016-06-22  6:49 ` Jan Beulich
  2016-06-23  9:56 ` David Vrabel
  1 sibling, 0 replies; 8+ messages in thread
From: Jan Beulich @ 2016-06-22  6:49 UTC (permalink / raw)
  To: Andrey Grodzovsky; +Cc: xen-devel, Boris Ostrovsky, stable, david.vrabel, jw

>>> On 21.06.16 at 20:26, <andrey2805@gmail.com> wrote:
> Current overlap check is evaluating to false a case where a filter field
> is fully contained (proper subset) of a r/w request.
> This change applies classical overlap check instead to include
> all the scenarios.
> 
> More specifically, for (Hilscher GmbH CIFX 50E-DP(M/S)) device
> driver the logic is such that the entire confspace  is read and
> written in 4 byte chunks.In this case as an example, CACHE_LINE_SIZE,
> LATENCY_TIMER and PCI_BIST are arriving together in one call to
> xen_pcibk_config_write with offset == 0xc and size == 4.
> With the exsisting overlap check LATENCY_TIMER field
> (offset == 0xd, length == 1) is fully contained in the write request
> and hence is excluded from write, which is incorrect.
> 
> Related to https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html 
> 
> Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> Reviewed-by: David Vrabel <david.vrabel@citrix.com>

Reviewed-by: Jan Beulich <jbeulich@suse.com>


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH] xen/pciback: Fix conf_space read/write overlap check.
@ 2016-06-21 18:26 Andrey Grodzovsky
  0 siblings, 0 replies; 8+ messages in thread
From: Andrey Grodzovsky @ 2016-06-21 18:26 UTC (permalink / raw)
  To: xen-devel
  Cc: jw, stable, david.vrabel, Jan Beulich, Boris Ostrovsky,
	Andrey Grodzovsky

Current overlap check is evaluating to false a case where a filter field
is fully contained (proper subset) of a r/w request.
This change applies classical overlap check instead to include
all the scenarios.

More specifically, for (Hilscher GmbH CIFX 50E-DP(M/S)) device
driver the logic is such that the entire confspace  is read and
written in 4 byte chunks.In this case as an example, CACHE_LINE_SIZE,
LATENCY_TIMER and PCI_BIST are arriving together in one call to
xen_pcibk_config_write with offset == 0xc and size == 4.
With the exsisting overlap check LATENCY_TIMER field
(offset == 0xd, length == 1) is fully contained in the write request
and hence is excluded from write, which is incorrect.

Related to https://www.mail-archive.com/xen-devel@lists.xen.org/msg72174.html

Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: David Vrabel <david.vrabel@citrix.com>
Cc: Jan Beulich <JBeulich@suse.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: stable@vger.kernel.org
Signed-off-by: Andrey Grodzovsky <andrey2805@gmail.com>
---
 drivers/xen/xen-pciback/conf_space.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c
index 8e67336..6a25533 100644
--- a/drivers/xen/xen-pciback/conf_space.c
+++ b/drivers/xen/xen-pciback/conf_space.c
@@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int offset, int size,
 		field_start = OFFSET(cfg_entry);
 		field_end = OFFSET(cfg_entry) + field->size;
 
-		if ((req_start >= field_start && req_start < field_end)
-		    || (req_end > field_start && req_end <= field_end)) {
+		 if (req_end > field_start && field_end > req_start) {
 			err = conf_space_read(dev, cfg_entry, field_start,
 					      &tmp_val);
 			if (err)
@@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value)
 		field_start = OFFSET(cfg_entry);
 		field_end = OFFSET(cfg_entry) + field->size;
 
-		if ((req_start >= field_start && req_start < field_end)
-		    || (req_end > field_start && req_end <= field_end)) {
+		 if (req_end > field_start && field_end > req_start) {
 			tmp_val = 0;
 
 			err = xen_pcibk_config_read(dev, field_start,
-- 
2.1.4


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply related	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2016-06-23  9:56 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-06-21 14:37 [PATCH] xen/pciback: Fix conf_space read/write overlap check Andrey Grodzovsky
2016-06-21 15:52 ` Boris Ostrovsky
2016-06-22  4:11   ` Juergen Gross
2016-06-21 17:15 ` David Vrabel
     [not found] ` <57697614.5020003@citrix.com>
2016-06-21 18:20   ` Andrey Grodzovsky
2016-06-21 18:26 Andrey Grodzovsky
     [not found] <1466533596-15122-1-git-send-email-andrey2805@gmail.com>
2016-06-22  6:49 ` Jan Beulich
2016-06-23  9:56 ` David Vrabel

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).