Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-09 Thread Neil Horman
On Mon, Nov 09, 2015 at 10:35:42PM +0200, Julian Anastasov wrote: > > Hello, > > On Sun, 8 Nov 2015, Neil Horman wrote: > > > On Sat, Nov 07, 2015 at 01:49:25AM +0200, Julian Anastasov wrote: > > > > > > flush can provide many parameters. As there is no > > > any kind of indication in t

Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-09 Thread Julian Anastasov
Hello, On Sun, 8 Nov 2015, Neil Horman wrote: > On Sat, Nov 07, 2015 at 01:49:25AM +0200, Julian Anastasov wrote: > > > > flush can provide many parameters. As there is no > > any kind of indication in the netlink message that all addresses > > are removed, we can not avoid the prom

Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-09 Thread Neil Horman
On Mon, Nov 09, 2015 at 07:31:38AM +0900, David Miller wrote: > From: Neil Horman > Date: Sun, 8 Nov 2015 16:56:41 -0500 > > > All I'm doing is, in effect masking the promote_secondaries sysctl > > for an interface doing a flush operation. > > Packets and other entities not controlled by RTNL ca

Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-08 Thread David Miller
From: Neil Horman Date: Sun, 8 Nov 2015 16:56:41 -0500 > All I'm doing is, in effect masking the promote_secondaries sysctl > for an interface doing a flush operation. Packets and other entities not controlled by RTNL can see an inconsistent state: no primary address on the interface. You can a

Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-08 Thread Neil Horman
On Sat, Nov 07, 2015 at 01:49:25AM +0200, Julian Anastasov wrote: > > Hello, > > On Fri, 6 Nov 2015, Neil Horman wrote: > > > The solution is to recognize that its pointless to promote an address to be > > a > > new primary, if there is a possibility that it will just be removed in the >

Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-07 Thread David Miller
From: Julian Anastasov Date: Sat, 7 Nov 2015 01:49:25 +0200 (EET) > On Fri, 6 Nov 2015, Neil Horman wrote: > >> The solution is to recognize that its pointless to promote an address to be a >> new primary, if there is a possibility that it will just be removed in the >> near >> future. As such

Re: [PATCH] inet: delay address promotion check until last request in message

2015-11-06 Thread Julian Anastasov
Hello, On Fri, 6 Nov 2015, Neil Horman wrote: > The solution is to recognize that its pointless to promote an address to be a > new primary, if there is a possibility that it will just be removed in the > near > future. As such this patch peeks ahead to the next request in the provided

[PATCH] inet: delay address promotion check until last request in message

2015-11-06 Thread Neil Horman
I recently got a report that, after a user added 40,000 addresses to an interface and ran ip addr flush , their system blocked on interface access for literally hours, and issued hundreds of hung task warnings. due to rtnl being held during this operation. While its certainly arguable that long de