Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[PATCH 1/2] ixgbe: define IXGBE_MAX_VFS_DRV_LIMIT macro and cleanup const 63

7 views
Skip to first unread message

Ethan Zhao

unread,
Dec 24, 2013, 11:10:02 AM12/24/13
to
Because ixgbe driver limit the max number of VF functions could be enalbed
to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the const 63
in code.

Signed-off-by: Ethan Zhao <ethan....@gmail.com>
---
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 4 ++--
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 5 +++--
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h | 5 +++++
3 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index 0ade0cd..47e9b44 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -4818,7 +4818,7 @@ static int ixgbe_sw_init(struct ixgbe_adapter *adapter)
#ifdef CONFIG_PCI_IOV
/* assign number of SR-IOV VFs */
if (hw->mac.type != ixgbe_mac_82598EB)
- adapter->num_vfs = (max_vfs > 63) ? 0 : max_vfs;
+ adapter->num_vfs = (max_vfs > IXGBE_MAX_VFS_DRV_LIMIT) ? 0 : max_vfs;

#endif
/* enable itr by default in dynamic mode */
@@ -7640,7 +7640,7 @@ static int ixgbe_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
ixgbe_init_mbx_params_pf(hw);
memcpy(&hw->mbx.ops, ii->mbx_ops, sizeof(hw->mbx.ops));
ixgbe_enable_sriov(adapter);
- pci_sriov_set_totalvfs(pdev, 63);
+ pci_sriov_set_totalvfs(pdev, IXGBE_MAX_VFS_DRV_LIMIT);
skip_sriov:

#endif
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
index 276d7b1..6925979 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
@@ -152,7 +152,8 @@ void ixgbe_enable_sriov(struct ixgbe_adapter *adapter)
* physical function. If the user requests greater thn
* 63 VFs then it is an error - reset to default of zero.
*/
- adapter->num_vfs = min_t(unsigned int, adapter->num_vfs, 63);
+ adapter->num_vfs = min_t(unsigned int,
+ adapter->num_vfs, IXGBE_MAX_VFS_DRV_LIMIT);

err = pci_enable_sriov(adapter->pdev, adapter->num_vfs);
if (err) {
@@ -259,7 +260,7 @@ static int ixgbe_pci_sriov_enable(struct pci_dev *dev, int num_vfs)
* PF. The PCI bus driver already checks for other values out of
* range.
*/
- if (num_vfs > 63) {
+ if (num_vfs > IXGBE_MAX_VFS_DRV_LIMIT) {
err = -EPERM;
goto err_out;
}
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
index 4713f9f..2593666 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
@@ -28,6 +28,11 @@
#ifndef _IXGBE_SRIOV_H_
#define _IXGBE_SRIOV_H_

+/* ixgbe driver limit the max number of VFs could be enabled to
+ * 63 (IXGBE_MAX_VF_FUNCTIONS－1)
+ */
+#define IXGBE_MAX_VFS_DRV_LIMIT (IXGBE_MAX_VF_FUNCTIONS－1)
+
void ixgbe_restore_vf_multicasts(struct ixgbe_adapter *adapter);
void ixgbe_msg_task(struct ixgbe_adapter *adapter);
int ixgbe_vf_configuration(struct pci_dev *pdev, unsigned int event_mask);
--
1.8.3.4 (Apple Git-47)

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Ethan Zhao

unread,
Dec 24, 2013, 11:20:01 AM12/24/13
to
Because ixgbe driver limit the max number of VF functions could be enabled
to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the const 63
in code.

v2: fix a typo.

Jeff Kirsher

unread,
Dec 27, 2013, 4:10:02 AM12/27/13
to
On Wed, 2013-12-25 at 00:12 +0800, Ethan Zhao wrote:
> Because ixgbe driver limit the max number of VF functions could be
> enabled
> to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the
> const 63
> in code.
>
> v2: fix a typo.
>
> Signed-off-by: Ethan Zhao <ethan....@gmail.com>
> ---
> drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 4 ++--
> drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 5 +++--
> drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h | 5 +++++
> 3 files changed, 10 insertions(+), 4 deletions(-)

Added to my queue, thanks Ethan!
signature.asc

Brown, Aaron F

unread,
Jan 14, 2014, 10:50:02 PM1/14/14
to

Hi Ethan,

Did Jeff contact you about this failing to compile? I'm currently
providing vacation covering for him and we found this was failing to
compile just before he left. We captured the failure in our notes for
this but there is no comment on if you were contacted or not.

Regardless, when I apply this patch (with or without 2-2) we get the
following error on a compilation attempt: Here's the error:
--------------------------------------------------------
Here's the error:
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c: In function
"ixgbe_sw_init":
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:5033: error: stray "\357"
in program
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:5033: error: stray "\274"
in program
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:5033: error: stray "\215"
in program
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:5033: error: expected ")"
before numeric constant
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c: In function
"ixgbe_probe":
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:7977: error: stray "\357"
in program
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:7977: error: stray "\274"
in program
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:7977: error: stray "\215"
in program
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c:7977: error: expected ")"
before numeric constant
make[5]: *** [drivers/net/ethernet/intel/ixgbe/ixgbe_main.o] Error 1
make[5]: *** Waiting for unfinished jobs....
make[4]: *** [drivers/net/ethernet/intel/ixgbe] Error 2
make[4]: *** Waiting for unfinished jobs....
make[3]: *** [drivers/net/ethernet/intel] Error 2
make[2]: *** [drivers/net/ethernet] Error 2
make[1]: *** [drivers/net] Error 2
make: *** [drivers] Error 2
--------------------------------------------------------

Thanks,
Aaron

Ethan Zhao

unread,
Jan 15, 2014, 8:10:02 AM1/15/14
to
Aaron,
I will check the patch and make it pass the building, seems encoding issue.

Thanks,
Ethan

Ethan Zhao

unread,
Jan 15, 2014, 9:20:02 AM1/15/14
to
Because ixgbe driver limit the max number of VF functions could be enabled
to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the const 63
in code.

v2: fix a typo.
v3: fix a encoding issue.

Signed-off-by: Ethan Zhao <ethan....@gmail.com>
---
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 4 ++--
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 5 +++--
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h | 5 +++++
3 files changed, 10 insertions(+), 4 deletions(-)

+ * 63 (IXGBE_MAX_VF_FUNCTIONS - 1)
+ */
+#define IXGBE_MAX_VFS_DRV_LIMIT (IXGBE_MAX_VF_FUNCTIONS - 1)
+
void ixgbe_restore_vf_multicasts(struct ixgbe_adapter *adapter);
void ixgbe_msg_task(struct ixgbe_adapter *adapter);
int ixgbe_vf_configuration(struct pci_dev *pdev, unsigned int event_mask);
--
1.8.3.4 (Apple Git-47)

Brown, Aaron F

unread,
Jan 15, 2014, 5:10:05 PM1/15/14
to
On Wed, 2014-01-15 at 22:12 +0800, Ethan Zhao wrote:
> Because ixgbe driver limit the max number of VF functions could be enabled
> to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the const 63
> in code.
>
> v2: fix a typo.
> v3: fix a encoding issue.
>
> Signed-off-by: Ethan Zhao <ethan....@gmail.com>
> ---
> drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 4 ++--
> drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 5 +++--
> drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h | 5 +++++
> 3 files changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
> index 0ade0cd..47e9b44 100644
> --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
> +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
> @@ -4818,7 +4818,7 @@ static int ixgbe_sw_init(struct ixgbe_adapter *adapter)
> #ifdef CONFIG_PCI_IOV
> /* assign number of SR-IOV VFs */
> if (hw->mac.type != ixgbe_mac_82598EB)
> - adapter->num_vfs = (max_vfs > 63) ? 0 : max_vfs;


Unfortunately the if statement got changed considerably with a recent
commit:

commit 170e85430bcbe4d18e81b5a70bb163c741381092
ixgbe: add warning when max_vfs is out of range.

And the pattern no longer exists to make a match. In other words, this
patch no longer applies to net-next and I have to ask you for yet
another spin if you still want to squash the magic number.

Thanks,
Aaron

Ethan Zhao

unread,
Jan 15, 2014, 8:30:02 PM1/15/14
to
It's not a good news. Our distro is waiting for this patch showing up in stable.
OK, info me if there is a windows for me to revise the patch.

Thanks,
Ethan

>
> Thanks,
> Aaron

Ethan Zhao

unread,
Jan 15, 2014, 9:00:02 PM1/15/14
to
Aaron,

Is this your net-next repo ? if so, I rebuild the patch with this repo
right now .
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next.git

Thanks,
Ethan

On Thu, Jan 16, 2014 at 9:54 AM, Brown, Aaron F <aaron....@intel.com> wrote:
> I don't think any particular window of time is better than another. I
> don't see this change as needing very thorough testing so if you send in
> (yet) another version I'll try to get it through our internal process as
> rapidly as I can.

Brown, Aaron F

unread,
Jan 15, 2014, 9:00:02 PM1/15/14
to
On Thu, 2014-01-16 at 09:27 +0800, Ethan Zhao wrote:

I don't think any particular window of time is better than another. I

Brown, Aaron F

unread,
Jan 15, 2014, 10:00:01 PM1/15/14
to
On Thu, 2014-01-16 at 09:58 +0800, Ethan Zhao wrote:
> Aaron,
>
> Is this your net-next repo ? if so, I rebuild the patch with this repo
> right now .
> git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next.git
>
> Thanks,
> Ethan
>
Only sort of. Jeff uses it to push patches up, but I don't have an
account there so am simply sending them up the old fashioned way, via
email, hence it is probably not getting updated while Jeff is out.

As long as your patch can apply cleanly to Dave Miller's net-next tree I
should be able to pull it into our internal ones.

Ethan Zhao

unread,
Jan 15, 2014, 10:10:01 PM1/15/14
to
Aaron,
Ok, Dave Miler's net-next tree.

Thanks,
Etan

ethan zhao

unread,
Jan 15, 2014, 11:30:02 PM1/15/14
to

Because ixgbe driver limit the max number of VF functions could be enalbed
to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the const 63
in code.

v3: revised for net-next tree.

Signed-off-by: Ethan Zhao <ethan...@oracle.com>
Signed-off-by: Ethan Zhao <ethan....@gmail.com>
---
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 4 ++--
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 4 ++--
drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h | 5 +++++
3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index cc06854..bea2cec 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -5028,7 +5028,7 @@ static int ixgbe_sw_init(struct ixgbe_adapter
*adapter)

/* assign number of SR-IOV VFs */
if (hw->mac.type != ixgbe_mac_82598EB) {
- if (max_vfs > 63) {
+ if (max_vfs > IXGBE_MAX_VFS_DRV_LIMIT) {
adapter->num_vfs = 0;
e_dev_warn("max_vfs parameter out of range. Not assigning
any SR-IOV VFs\n");
} else {
@@ -7973,7 +7973,7 @@ static int ixgbe_probe(struct pci_dev *pdev, const
struct pci_device_id *ent)
ixgbe_init_mbx_params_pf(hw);
memcpy(&hw->mbx.ops, ii->mbx_ops, sizeof(hw->mbx.ops));
ixgbe_enable_sriov(adapter);
- pci_sriov_set_totalvfs(pdev, 63);
+ pci_sriov_set_totalvfs(pdev, IXGBE_MAX_VFS_DRV_LIMIT);
skip_sriov:

#endif
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
index 359f6e6..b324260 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c
@@ -148,7 +148,7 @@ void ixgbe_enable_sriov(struct ixgbe_adapter *adapter)
* physical function. If the user requests greater thn
* 63 VFs then it is an error - reset to default of zero.
*/
- adapter->num_vfs = min_t(unsigned int, adapter->num_vfs, 63);
+ adapter->num_vfs = min_t(unsigned int, adapter->num_vfs,
IXGBE_MAX_VFS_DRV_LIMIT);

err = pci_enable_sriov(adapter->pdev, adapter->num_vfs);
if (err) {
@@ -257,7 +257,7 @@ static int ixgbe_pci_sriov_enable(struct pci_dev
*dev, int num_vfs)
* PF. The PCI bus driver already checks for other values out of
* range.
*/
- if (num_vfs > 63) {
+ if (num_vfs > IXGBE_MAX_VFS_DRV_LIMIT) {
err = -EPERM;
goto err_out;
}
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
index 4713f9f..8bd2919 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.h
@@ -28,6 +28,11 @@
#ifndef _IXGBE_SRIOV_H_
#define _IXGBE_SRIOV_H_

+/* ixgbe driver limit the max number of VFs could be enabled to
+ * 63 (IXGBE_MAX_VF_FUNCTIONS - 1)
+ */
+#define IXGBE_MAX_VFS_DRV_LIMIT (IXGBE_MAX_VF_FUNCTIONS - 1)
+
void ixgbe_restore_vf_multicasts(struct ixgbe_adapter *adapter);
void ixgbe_msg_task(struct ixgbe_adapter *adapter);
int ixgbe_vf_configuration(struct pci_dev *pdev, unsigned int event_mask);
--
1.7.1

Ethan Zhao

unread,
Jan 15, 2014, 11:30:02 PM1/15/14
to
Aaron,
Revised those patches for Dave Miller's net-next OK, passed
building. resent.

Thanks,
Ethan

David Miller

unread,
Jan 16, 2014, 1:00:03 AM1/16/14
to
From: ethan zhao <ethan...@oracle.com>
Date: Thu, 16 Jan 2014 12:25:01 +0800

> Signed-off-by: Ethan Zhao <ethan...@oracle.com>
> Signed-off-by: Ethan Zhao <ethan....@gmail.com>

Please don't give two signoffs for yourself, it is not appropriate at
all.

I am very genuinely curious where you got the idea to do that,
particularly as I've never seen anyone else do it before. So it's
really not possible that you got the idea from someone else's actions.

Ethan Zhao

unread,
Jan 16, 2014, 1:20:01 AM1/16/14
to
David,
Sorry about the signoffs, I am confused about how to sign the patch,
If I sign it with off-work mail address, some of it is done within work-hours,
maybe that why I sign if with both personal mail address and work one.
If that bother you, just remove it.

Thanks,
Ethan

ethan zhao

unread,
Jan 16, 2014, 1:20:02 AM1/16/14
to

Because ixgbe driver limit the max number of VF functions could be enalbed
to 63, so define one macro IXGBE_MAX_VFS_DRV_LIMIT and cleanup the const 63
in code.

v3: revised for net-next tree.
V4: remove one signoff of two.
0 new messages