Message ID | 1574451706-19058-1-git-send-email-tlfalcon@linux.ibm.com |
---|---|
Headers | show |
Series | ibmvnic: Harden device commands and queries | expand |
On Fri, 22 Nov 2019 13:41:42 -0600, Thomas Falcon wrote: > This patch series fixes some shortcomings with the current > VNIC device command implementation. The first patch fixes > the initialization of driver completion structures used > for device commands. Additionally, all waits for device > commands are bounded with a timeout in the event that the > device does not respond or becomes inoperable. Finally, > serialize queries to retain the integrity of device return > codes. I have minor comments on two patches, but also I think it's a little late in the release cycle for putting this in net. Could you target net-next and repost ASAP so it still makes it into 5.5? Thanks.
On 11/23/19 7:49 PM, Jakub Kicinski wrote: > On Fri, 22 Nov 2019 13:41:42 -0600, Thomas Falcon wrote: >> This patch series fixes some shortcomings with the current >> VNIC device command implementation. The first patch fixes >> the initialization of driver completion structures used >> for device commands. Additionally, all waits for device >> commands are bounded with a timeout in the event that the >> device does not respond or becomes inoperable. Finally, >> serialize queries to retain the integrity of device return >> codes. > I have minor comments on two patches, but also I think it's > a little late in the release cycle for putting this in net. > > Could you target net-next and repost ASAP so it still makes > it into 5.5? > > Thanks. Thank you, sorry for the late response. I will make the requested changes ASAP, but I've missed the net-next window. What should I target for v2? Thanks again, Tom
On Mon, 25 Nov 2019 12:40:42 -0600, Thomas Falcon wrote: > On 11/23/19 7:49 PM, Jakub Kicinski wrote: > > On Fri, 22 Nov 2019 13:41:42 -0600, Thomas Falcon wrote: > >> This patch series fixes some shortcomings with the current > >> VNIC device command implementation. The first patch fixes > >> the initialization of driver completion structures used > >> for device commands. Additionally, all waits for device > >> commands are bounded with a timeout in the event that the > >> device does not respond or becomes inoperable. Finally, > >> serialize queries to retain the integrity of device return > >> codes. > > I have minor comments on two patches, but also I think it's > > a little late in the release cycle for putting this in net. > > > > Could you target net-next and repost ASAP so it still makes > > it into 5.5? > > Thank you, sorry for the late response. I will make the requested > changes ASAP, but I've missed the net-next window. What should I target > for v2? You're right, sticking to "net" makes sense at this point.