kernel.git/drivers/nvme/host/ioctl.c, branch linux-5.19.y

nvme: handle effects after freeing the request

2022-10-24T07:58:28Z

[ Upstream commit bc8fb906b0ff9339b4286698cb7cd9cd5b8c53eb ] If a reset occurs after the scan work attempts to issue a command, the reset may quisce the admin queue, which blocks the scan work's command from dispatching. The scan work will not be able to complete while the queue is quiesced. Meanwhile, the reset work will cancel all outstanding admin tags and wait until all requests have transitioned to idle, which includes the passthrough request. But the passthrough request won't be set to idle until after the scan_work flushes, so we're deadlocked. Fix this by handling the end effects after the request has been freed. Link: https://bugzilla.kernel.org/show_bug.cgi?id=216354 Reported-by: Jonathan Derrick Signed-off-by: Keith Busch Reviewed-by: Sagi Grimberg Reviewed-by: Chao Leng Signed-off-by: Christoph Hellwig Signed-off-by: Sasha Levin

blk-mq: remove the done argument to blk_execute_rq_nowait

2022-05-28T12:15:27Z

Let the caller set it together with the end_io_data instead of passing a pointless argument. Note the the target code did in fact already set it and then just overrode it again by calling blk_execute_rq_nowait. Signed-off-by: Christoph Hellwig Reviewed-by: Keith Busch Reviewed-by: Kanchan Joshi Reviewed-by: Chaitanya Kulkarni Link: https://lore.kernel.org/r/20220524121530.943123-4-hch@lst.de Signed-off-by: Jens Axboe

nvme: enable uring-passthrough for admin commands

2022-05-20T12:17:33Z

Add two new opcodes that userspace can use for admin commands: NVME_URING_CMD_ADMIN : non-vectroed NVME_URING_CMD_ADMIN_VEC : vectored variant Wire up support when these are issued on controller node(/dev/nvmeX). Signed-off-by: Kanchan Joshi Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20220520090630.70394-3-joshi.k@samsung.com Signed-off-by: Jens Axboe

nvme: helper for uring-passthrough checks

2022-05-20T12:17:33Z

Factor out a helper consolidating the error checks, and fix typo in a comment too. This is in preparation to support admin commands on this path. Signed-off-by: Kanchan Joshi Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20220520090630.70394-2-joshi.k@samsung.com Signed-off-by: Jens Axboe

nvme: add vectored-io support for uring-cmd

2022-05-11T13:41:19Z

wire up support for async passthru that takes an array of buffers (using iovec). Exposed via a new op NVME_URING_CMD_IO_VEC. Same 'struct nvme_uring_cmd' is to be used with - 1. cmd.addr as base address of user iovec array 2. cmd.data_len as count of iovec array elements Signed-off-by: Kanchan Joshi Signed-off-by: Anuj Gupta Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20220511054750.20432-6-joshi.k@samsung.com Signed-off-by: Jens Axboe

nvme: wire-up uring-cmd support for io-passthru on char-device.

2022-05-11T13:41:13Z

Introduce handler for fops->uring_cmd(), implementing async passthru on char device (/dev/ngX). The handler supports newly introduced operation NVME_URING_CMD_IO. This operates on a new structure nvme_uring_cmd, which is similar to struct nvme_passthru_cmd64 but without the embedded 8b result field. This field is not needed since uring-cmd allows to return additional result via big-CQE. Signed-off-by: Kanchan Joshi Signed-off-by: Anuj Gupta Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20220511054750.20432-5-joshi.k@samsung.com Signed-off-by: Jens Axboe

nvme: refactor nvme_submit_user_cmd()

2022-05-11T13:41:05Z

Divide the work into two helpers, namely nvme_alloc_user_request and nvme_execute_user_rq. This is a prep patch, to help wiring up uring-cmd support in nvme. Signed-off-by: Christoph Hellwig [axboe: fold in fix for assuming bio is non-NULL] Link: https://lore.kernel.org/r/20220511054750.20432-4-joshi.k@samsung.com Signed-off-by: Jens Axboe

nvme: remove nvme_alloc_request and nvme_alloc_request_qid

2022-03-16T08:47:05Z

Just open code the allocation + initialization in the callers. Signed-off-by: Christoph Hellwig Reviewed-by: Keith Busch Reviewed-by: Sagi Grimberg Reviewed-by: Chaitanya Kulkarni

nvme: add vectored-io support for user-passthrough

2022-02-28T11:45:06Z

Add a new NVME_IOCTL_IO64_CMD_VEC ioctl that works like the existing NVME_IOCTL_IO64_CMD ioctl except that it takes and array of iovecs and thus supports vectored I/O. - cmd.addr is base address of user iovec array - cmd.vec_cnt is count of iovec array elements This patch does not include vectored-variant for admin-commands as most of them are light on buffers and likely to have low invocation frequency. Signed-off-by: Kanchan Joshi Reviewed-by: Sagi Grimberg Signed-off-by: Christoph Hellwig

remove the lightnvm subsystem

2021-08-14T21:54:09Z

Lightnvm supports the OCSSD 1.x and 2.0 specs which were early attempts to produce Open Channel SSDs and never made it into the NVMe spec proper. They have since been superceeded by NVMe enhancements such as ZNS support. Remove the support per the deprecation schedule. Signed-off-by: Christoph Hellwig Link: https://lore.kernel.org/r/20210812132308.38486-1-hch@lst.de Reviewed-by: Matias Bjørling Reviewed-by: Javier González Signed-off-by: Jens Axboe