freebsd-nq/sys/dev/nvd/nvd.c
Jim Harris 70fb74bd12 nvd: set d_delmaxsize to full capacity of NVMe namespace
The NVMe specification has no ability to specify a maximum delete size
that is less than the full capacity of the namespace - so just using the
namespace size is the correct value here.

This fixes reported issues where ZFS trim on init looked like it was
hanging the system - previously the default I/O max size (128KB on
Intel NVMe controllers) was used for delete operations which worked out
to only about 8MB/s.  With this patch I can add an 800GB DC P3700
drive to a ZFS pool in about 15-20 seconds.

Reported by: Dylan Just <dylan@techtangents.com>
MFC after:	3 days
Sponsored by:	Intel
2015-07-21 20:53:21 +00:00

394 lines
9.0 KiB
C

/*-
* Copyright (C) 2012-2013 Intel Corporation
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/bio.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/module.h>
#include <sys/systm.h>
#include <sys/taskqueue.h>
#include <geom/geom.h>
#include <geom/geom_disk.h>
#include <dev/nvme/nvme.h>
#define NVD_STR "nvd"
struct nvd_disk;
static disk_ioctl_t nvd_ioctl;
static disk_strategy_t nvd_strategy;
static void *nvd_new_disk(struct nvme_namespace *ns, void *ctrlr);
static void destroy_geom_disk(struct nvd_disk *ndisk);
static void *nvd_new_controller(struct nvme_controller *ctrlr);
static void nvd_controller_fail(void *ctrlr);
static int nvd_load(void);
static void nvd_unload(void);
MALLOC_DEFINE(M_NVD, "nvd", "nvd(4) allocations");
struct nvme_consumer *consumer_handle;
struct nvd_disk {
struct bio_queue_head bioq;
struct task bioqtask;
struct mtx bioqlock;
struct disk *disk;
struct taskqueue *tq;
struct nvme_namespace *ns;
uint32_t cur_depth;
TAILQ_ENTRY(nvd_disk) global_tailq;
TAILQ_ENTRY(nvd_disk) ctrlr_tailq;
};
struct nvd_controller {
TAILQ_ENTRY(nvd_controller) tailq;
TAILQ_HEAD(, nvd_disk) disk_head;
};
static TAILQ_HEAD(, nvd_controller) ctrlr_head;
static TAILQ_HEAD(disk_list, nvd_disk) disk_head;
static int nvd_modevent(module_t mod, int type, void *arg)
{
int error = 0;
switch (type) {
case MOD_LOAD:
error = nvd_load();
break;
case MOD_UNLOAD:
nvd_unload();
break;
default:
break;
}
return (error);
}
moduledata_t nvd_mod = {
NVD_STR,
(modeventhand_t)nvd_modevent,
0
};
DECLARE_MODULE(nvd, nvd_mod, SI_SUB_DRIVERS, SI_ORDER_ANY);
MODULE_VERSION(nvd, 1);
MODULE_DEPEND(nvd, nvme, 1, 1, 1);
static int
nvd_load()
{
TAILQ_INIT(&ctrlr_head);
TAILQ_INIT(&disk_head);
consumer_handle = nvme_register_consumer(nvd_new_disk,
nvd_new_controller, NULL, nvd_controller_fail);
return (consumer_handle != NULL ? 0 : -1);
}
static void
nvd_unload()
{
struct nvd_controller *ctrlr;
struct nvd_disk *disk;
while (!TAILQ_EMPTY(&ctrlr_head)) {
ctrlr = TAILQ_FIRST(&ctrlr_head);
TAILQ_REMOVE(&ctrlr_head, ctrlr, tailq);
free(ctrlr, M_NVD);
}
while (!TAILQ_EMPTY(&disk_head)) {
disk = TAILQ_FIRST(&disk_head);
TAILQ_REMOVE(&disk_head, disk, global_tailq);
destroy_geom_disk(disk);
free(disk, M_NVD);
}
nvme_unregister_consumer(consumer_handle);
}
static void
nvd_strategy(struct bio *bp)
{
struct nvd_disk *ndisk;
ndisk = (struct nvd_disk *)bp->bio_disk->d_drv1;
mtx_lock(&ndisk->bioqlock);
bioq_insert_tail(&ndisk->bioq, bp);
mtx_unlock(&ndisk->bioqlock);
taskqueue_enqueue(ndisk->tq, &ndisk->bioqtask);
}
static int
nvd_ioctl(struct disk *ndisk, u_long cmd, void *data, int fflag,
struct thread *td)
{
int ret = 0;
switch (cmd) {
default:
ret = EIO;
}
return (ret);
}
static void
nvd_done(void *arg, const struct nvme_completion *cpl)
{
struct bio *bp;
struct nvd_disk *ndisk;
bp = (struct bio *)arg;
ndisk = bp->bio_disk->d_drv1;
atomic_add_int(&ndisk->cur_depth, -1);
biodone(bp);
}
static void
nvd_bioq_process(void *arg, int pending)
{
struct nvd_disk *ndisk = arg;
struct bio *bp;
int err;
for (;;) {
mtx_lock(&ndisk->bioqlock);
bp = bioq_takefirst(&ndisk->bioq);
mtx_unlock(&ndisk->bioqlock);
if (bp == NULL)
break;
#ifdef BIO_ORDERED
/*
* BIO_ORDERED flag dictates that all outstanding bios
* must be completed before processing the bio with
* BIO_ORDERED flag set.
*/
if (bp->bio_flags & BIO_ORDERED) {
while (ndisk->cur_depth > 0) {
pause("nvd flush", 1);
}
}
#endif
bp->bio_driver1 = NULL;
atomic_add_int(&ndisk->cur_depth, 1);
err = nvme_ns_bio_process(ndisk->ns, bp, nvd_done);
if (err) {
atomic_add_int(&ndisk->cur_depth, -1);
bp->bio_error = err;
bp->bio_flags |= BIO_ERROR;
bp->bio_resid = bp->bio_bcount;
biodone(bp);
}
#ifdef BIO_ORDERED
/*
* BIO_ORDERED flag dictates that the bio with BIO_ORDERED
* flag set must be completed before proceeding with
* additional bios.
*/
if (bp->bio_flags & BIO_ORDERED) {
while (ndisk->cur_depth > 0) {
pause("nvd flush", 1);
}
}
#endif
}
}
static void *
nvd_new_controller(struct nvme_controller *ctrlr)
{
struct nvd_controller *nvd_ctrlr;
nvd_ctrlr = malloc(sizeof(struct nvd_controller), M_NVD,
M_ZERO | M_WAITOK);
TAILQ_INIT(&nvd_ctrlr->disk_head);
TAILQ_INSERT_TAIL(&ctrlr_head, nvd_ctrlr, tailq);
return (nvd_ctrlr);
}
static void *
nvd_new_disk(struct nvme_namespace *ns, void *ctrlr_arg)
{
uint8_t descr[NVME_MODEL_NUMBER_LENGTH+1];
struct nvd_disk *ndisk;
struct disk *disk;
struct nvd_controller *ctrlr = ctrlr_arg;
ndisk = malloc(sizeof(struct nvd_disk), M_NVD, M_ZERO | M_WAITOK);
disk = disk_alloc();
disk->d_strategy = nvd_strategy;
disk->d_ioctl = nvd_ioctl;
disk->d_name = NVD_STR;
disk->d_drv1 = ndisk;
disk->d_maxsize = nvme_ns_get_max_io_xfer_size(ns);
disk->d_sectorsize = nvme_ns_get_sector_size(ns);
disk->d_mediasize = (off_t)nvme_ns_get_size(ns);
disk->d_delmaxsize = (off_t)nvme_ns_get_size(ns);
if (TAILQ_EMPTY(&disk_head))
disk->d_unit = 0;
else
disk->d_unit =
TAILQ_LAST(&disk_head, disk_list)->disk->d_unit + 1;
disk->d_flags = 0;
if (nvme_ns_get_flags(ns) & NVME_NS_DEALLOCATE_SUPPORTED)
disk->d_flags |= DISKFLAG_CANDELETE;
if (nvme_ns_get_flags(ns) & NVME_NS_FLUSH_SUPPORTED)
disk->d_flags |= DISKFLAG_CANFLUSHCACHE;
/* ifdef used here to ease porting to stable branches at a later point. */
#ifdef DISKFLAG_UNMAPPED_BIO
disk->d_flags |= DISKFLAG_UNMAPPED_BIO;
#endif
/*
* d_ident and d_descr are both far bigger than the length of either
* the serial or model number strings.
*/
nvme_strvis(disk->d_ident, nvme_ns_get_serial_number(ns),
sizeof(disk->d_ident), NVME_SERIAL_NUMBER_LENGTH);
nvme_strvis(descr, nvme_ns_get_model_number(ns), sizeof(descr),
NVME_MODEL_NUMBER_LENGTH);
#if __FreeBSD_version >= 900034
strlcpy(disk->d_descr, descr, sizeof(descr));
#endif
ndisk->ns = ns;
ndisk->disk = disk;
ndisk->cur_depth = 0;
mtx_init(&ndisk->bioqlock, "NVD bioq lock", NULL, MTX_DEF);
bioq_init(&ndisk->bioq);
TASK_INIT(&ndisk->bioqtask, 0, nvd_bioq_process, ndisk);
ndisk->tq = taskqueue_create("nvd_taskq", M_WAITOK,
taskqueue_thread_enqueue, &ndisk->tq);
taskqueue_start_threads(&ndisk->tq, 1, PI_DISK, "nvd taskq");
TAILQ_INSERT_TAIL(&disk_head, ndisk, global_tailq);
TAILQ_INSERT_TAIL(&ctrlr->disk_head, ndisk, ctrlr_tailq);
disk_create(disk, DISK_VERSION);
printf(NVD_STR"%u: <%s> NVMe namespace\n", disk->d_unit, descr);
printf(NVD_STR"%u: %juMB (%ju %u byte sectors)\n", disk->d_unit,
(uintmax_t)disk->d_mediasize / (1024*1024),
(uintmax_t)disk->d_mediasize / disk->d_sectorsize,
disk->d_sectorsize);
return (NULL);
}
static void
destroy_geom_disk(struct nvd_disk *ndisk)
{
struct bio *bp;
struct disk *disk;
uint32_t unit;
int cnt = 0;
disk = ndisk->disk;
unit = disk->d_unit;
taskqueue_free(ndisk->tq);
disk_destroy(ndisk->disk);
mtx_lock(&ndisk->bioqlock);
for (;;) {
bp = bioq_takefirst(&ndisk->bioq);
if (bp == NULL)
break;
bp->bio_error = EIO;
bp->bio_flags |= BIO_ERROR;
bp->bio_resid = bp->bio_bcount;
cnt++;
biodone(bp);
}
printf(NVD_STR"%u: lost device - %d outstanding\n", unit, cnt);
printf(NVD_STR"%u: removing device entry\n", unit);
mtx_unlock(&ndisk->bioqlock);
mtx_destroy(&ndisk->bioqlock);
}
static void
nvd_controller_fail(void *ctrlr_arg)
{
struct nvd_controller *ctrlr = ctrlr_arg;
struct nvd_disk *disk;
while (!TAILQ_EMPTY(&ctrlr->disk_head)) {
disk = TAILQ_FIRST(&ctrlr->disk_head);
TAILQ_REMOVE(&disk_head, disk, global_tailq);
TAILQ_REMOVE(&ctrlr->disk_head, disk, ctrlr_tailq);
destroy_geom_disk(disk);
free(disk, M_NVD);
}
TAILQ_REMOVE(&ctrlr_head, ctrlr, tailq);
free(ctrlr, M_NVD);
}