[next,S83-V5,2/9] i40e: avoid divide by zero

Message ID 20171218101525.75683-1-alice.michael@intel.com
State Under Review
Delegated to: Jeff Kirsher
Headers show
Series
  • [next,S83-V5,1/9] i40e/i40evf: Enable NVMUpdate to retrieve AdminQ and add preservation flags for NVM update
Related show

Commit Message

Alice Michael Dec. 18, 2017, 10:15 a.m.
From: Mitch Williams <mitch.a.williams@intel.com>

In some weird circumstances with DCB enabled, the firmware can fail to
configure the VSI, leaving us with zero traffic classes. Check for this
state when we configure RSS to avoid a panic.

Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

Comments

Bowers, AndrewX Dec. 21, 2017, 7:51 p.m. | #1
> -----Original Message-----
> From: Intel-wired-lan [mailto:intel-wired-lan-bounces@osuosl.org] On
> Behalf Of Alice Michael
> Sent: Monday, December 18, 2017 2:15 AM
> To: Michael, Alice <alice.michael@intel.com>; intel-wired-
> lan@lists.osuosl.org
> Subject: [Intel-wired-lan] [next PATCH S83-V5 2/9] i40e: avoid divide by zero
> 
> From: Mitch Williams <mitch.a.williams@intel.com>
> 
> In some weird circumstances with DCB enabled, the firmware can fail to
> configure the VSI, leaving us with zero traffic classes. Check for this state
> when we configure RSS to avoid a panic.
> 
> Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
> ---
>  drivers/net/ethernet/intel/i40e/i40e_main.c | 9 +++++++--
>  1 file changed, 7 insertions(+), 2 deletions(-)

Tested-by: Andrew Bowers <andrewx.bowers@intel.com>

Patch

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 4f74156..7a882d7 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -10839,8 +10839,13 @@  static int i40e_pf_config_rss(struct i40e_pf *pf)
 	/* Determine the RSS size of the VSI */
 	if (!vsi->rss_size) {
 		u16 qcount;
-
-		qcount = vsi->num_queue_pairs / vsi->tc_config.numtc;
+		/* If the firmware does something weird during VSI init, we
+		 * could end up with zero TCs. Check for that to avoid
+		 * divide-by-zero. It probably won't pass traffic, but it also
+		 * won't panic.
+		 */
+		qcount = vsi->num_queue_pairs /
+			 (vsi->tc_config.numtc ? vsi->tc_config.numtc : 1);
 		vsi->rss_size = min_t(int, pf->alloc_rss_size, qcount);
 	}
 	if (!vsi->rss_size)