Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*\[RFC\s+PATCH\s+net\-next\s+0\/4\]\s+ipvs\:\s+check\s+scheduler\s+callback\s+return\s+values\s*$/: 4 ]

Total 4 documents matching your query.

1. Re: [RFC PATCH net-next 0/4] ipvs: check scheduler callback return values (score: 1)
Author: Julian Anastasov <ja@xxxxxx>
Date: Tue, 24 Feb 2015 11:17:47 +0200 (EET)
Hello, The schedulers do not need RCU for list traversal if they implement their own lists. But if dest is returned from the schedule() method this dest is not freed while in RCU grace period. And th
/html/lvs-devel/2015-02/msg00051.html (14,522 bytes)

2. Re: [RFC PATCH net-next 0/4] ipvs: check scheduler callback return values (score: 1)
Author: Alex Gartrell <agartrell@xxxxxx>
Date: Mon, 23 Feb 2015 23:51:05 -0800
Hey Julian, Thank you for your quick and thoughtful response :) You can use GFP_KERNEL from add_dest method. That should definitely improve things for us. Can a new create_dest(svc, bool for_update)
/html/lvs-devel/2015-02/msg00050.html (14,097 bytes)

3. Re: [RFC PATCH net-next 0/4] ipvs: check scheduler callback return values (score: 1)
Author: Julian Anastasov <ja@xxxxxx>
Date: Tue, 24 Feb 2015 09:30:06 +0200 (EET)
Hello, You can use GFP_KERNEL from add_dest method. Can a new create_dest(svc, bool for_update) method solve such problem? Such modifications are serialized, so create_dest can attach new structure t
/html/lvs-devel/2015-02/msg00049.html (12,513 bytes)

4. [RFC PATCH net-next 0/4] ipvs: check scheduler callback return values (score: 1)
Author: Alex Gartrell <agartrell@xxxxxx>
Date: Mon, 23 Feb 2015 21:07:38 -0800
We have a custom scheduler that is failing some atomic allocations sometimes, and it was made pretty hard to discover by the fact that we more or less drop the return code on the floor. This is espec
/html/lvs-devel/2015-02/msg00047.html (10,600 bytes)


This search system is powered by Namazu