From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga04.intel.com ([192.55.52.120]:18569 "EHLO mga04.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751628AbcELRpO (ORCPT ); Thu, 12 May 2016 13:45:14 -0400 From: Jon Derrick To: linux-block@vger.kernel.org Cc: Jon Derrick , "Jens Axboe" , "Alexander Viro" , linux-fsdevel@vger.kernel.org, "Dan Williams" , "Jeff Moyer" , "Stephen Bates" , "Keith Busch" , linux-nvme@lists.infradead.org, "Christoph Hellwig" Subject: [RFC 0/2] Support for high-priority block device flag Date: Thu, 12 May 2016 11:43:04 -0600 Message-Id: <1463074986-3070-1-git-send-email-jonathan.derrick@intel.com> Sender: linux-block-owner@vger.kernel.org List-Id: linux-block@vger.kernel.org This set intends to recreate block polling (now HIPRI) behavior that was present in 4.5, where all IO on a queue could be selected to use block polling behavior. The set allows a block device file to subscribe to block polling on a block device granularity, rather than a per-queue granularity. There have been few-to-no arguments in support of the per-queue, always-poll functionality that 4.5 offered, moreso in favor of enabling polling on the entire block device (or indivual IOs as 4.6 offers). I've been made aware that streams and ioprio may supercede this functionality in the future, but I'm hoping this is an acceptable stopgap in the meantime. This set applies against 4.6-rc7 as well as Jens' for-4.7/core (I've also been made aware that it may not apply cleanly to 4.7 after several DAX changes) Jon Derrick (2): block: allow other bd i_node flags when DAX is disabled block: Introduce S_HIPRI inode flag block/ioctl.c | 33 +++++++++++++++++++++++++++++++++ fs/block_dev.c | 5 ++++- include/linux/fs.h | 2 ++ include/uapi/linux/fs.h | 2 ++ 4 files changed, 41 insertions(+), 1 deletion(-) -- 1.8.3.1 From mboxrd@z Thu Jan 1 00:00:00 1970 From: jonathan.derrick@intel.com (Jon Derrick) Date: Thu, 12 May 2016 11:43:04 -0600 Subject: [RFC 0/2] Support for high-priority block device flag Message-ID: <1463074986-3070-1-git-send-email-jonathan.derrick@intel.com> This set intends to recreate block polling (now HIPRI) behavior that was present in 4.5, where all IO on a queue could be selected to use block polling behavior. The set allows a block device file to subscribe to block polling on a block device granularity, rather than a per-queue granularity. There have been few-to-no arguments in support of the per-queue, always-poll functionality that 4.5 offered, moreso in favor of enabling polling on the entire block device (or indivual IOs as 4.6 offers). I've been made aware that streams and ioprio may supercede this functionality in the future, but I'm hoping this is an acceptable stopgap in the meantime. This set applies against 4.6-rc7 as well as Jens' for-4.7/core (I've also been made aware that it may not apply cleanly to 4.7 after several DAX changes) Jon Derrick (2): block: allow other bd i_node flags when DAX is disabled block: Introduce S_HIPRI inode flag block/ioctl.c | 33 +++++++++++++++++++++++++++++++++ fs/block_dev.c | 5 ++++- include/linux/fs.h | 2 ++ include/uapi/linux/fs.h | 2 ++ 4 files changed, 41 insertions(+), 1 deletion(-) -- 1.8.3.1