Hello, I am not familiar with Linux mailing, sorry for this.
I am writing this mail because I am not sure about commit 6c16fa957e84c8b640dadc4e0264ff0d2dae7aa3 added in 3.18.93.
That backport made the arp.h changes in struct __ipv4_neigh_lookup(), but compared with upstream commit, I think that 3 lines should be added in __ipv4_neigh_lookup_noref(), as I go through the context. Also the latter would be called unconditionally if former is called so it would cover more cases.
Could you please check this and let me know? Though it would 90% be my derp.
Thanks, Wang
On Fri, Feb 02, 2018 at 10:58:02AM +0800, Wang Han wrote:
Hello, I am not familiar with Linux mailing, sorry for this.
I am writing this mail because I am not sure about commit 6c16fa957e84c8b640dadc4e0264ff0d2dae7aa3 added in 3.18.93.
That backport made the arp.h changes in struct __ipv4_neigh_lookup(), but compared with upstream commit, I think that 3 lines should be added in __ipv4_neigh_lookup_noref(), as I go through the context. Also the latter would be called unconditionally if former is called so it would cover more cases.
Could you please check this and let me know? Though it would 90% be my derp.
You might be right, I did the backport of this by hand, and odds are I got it wrong. Thanks for checking this and finding the issue.
Could you send me a patch to fix this up so I can apply it to the next 3.18.y release and give you the proper credit for it?
thanks,
greg k-h
Hello Greg,
I have already submitted a patch to linux-stable here (https://www.spinics.net/lists/stable/msg213160.html), would like to see it on 3.18.94 queue.
Thanks,
Wang
2018-02-02 17:37 GMT+08:00 Greg KH gregkh@linuxfoundation.org:
On Fri, Feb 02, 2018 at 10:58:02AM +0800, Wang Han wrote:
Hello, I am not familiar with Linux mailing, sorry for this.
I am writing this mail because I am not sure about commit 6c16fa957e84c8b640dadc4e0264ff0d2dae7aa3 added in 3.18.93.
That backport made the arp.h changes in struct __ipv4_neigh_lookup(), but compared with upstream commit, I think that 3 lines should be added in __ipv4_neigh_lookup_noref(), as I go through the context. Also the latter would be called unconditionally if former is called so it would cover more cases.
Could you please check this and let me know? Though it would 90% be my derp.
You might be right, I did the backport of this by hand, and odds are I got it wrong. Thanks for checking this and finding the issue.
Could you send me a patch to fix this up so I can apply it to the next 3.18.y release and give you the proper credit for it?
thanks,
greg k-h
On Tue, Feb 06, 2018 at 09:17:56AM +0800, Wang Han wrote:
Hello Greg,
I have already submitted a patch to linux-stable here (https://www.spinics.net/lists/stable/msg213160.html), would like to see it on 3.18.94 queue.
Now applied, sorry for the delay.
greg k-h
linux-stable-mirror@lists.linaro.org