[V5] nvme: enable ro namespace for ZNS without append
javier at javigon.com
javier at javigon.com
Tue Dec 1 07:02:21 EST 2020
From: Javier González <javier.gonz at samsung.com>
Allow ZNS NVMe SSDs to present a read-only namespace when append is not
supported, instead of rejecting the namespace directly.
This allows (i) the namespace to be used in read-only mode, which is not
a problem as the append command only affects the write path, and (ii) to
use standard management tools such as nvme-cli to choose a different
format or firmware slot that is compatible with the Linux zoned block
device.
This patch includes comments from Christoph, Niklas and Keith that
applied to a different approach setting capacity to 0
https://www.spinics.net/lists/linux-block/msg60747.html
The reminder of the original patch will be submitted separately.
Changes since V4:
- Revert re-reading effects to page. Wait for patches from Christoph
on RO flag refactoring that allow to clear the flags.
Changes since V3:
- Force re-reading effects log page (from Keith)
- Pending proper refresh of RO flag - dependency on patch from
Christoph
Changes since V2:
- Fix small conflict with a queued patch from Sagi (from Christoph)
- Fix indentation (from Niklas)
- Refresh effects log page to account for FW changes (from Keith)
Changes since V1:
- Change logic to use NVME_NS_ATTR_RO (from Christoph)
- Set max_zone_append egen in RO. This allows the device to be
properly revalidated and enables user-space tools such as blkzone to
be used when interacting with this zoned device.
Signed-off-by: Javier González <javier.gonz at samsung.com>
---
drivers/nvme/host/core.c | 2 +-
drivers/nvme/host/nvme.h | 1 +
drivers/nvme/host/zns.c | 14 +++++++++-----
3 files changed, 11 insertions(+), 6 deletions(-)
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index f42ed266bd7f..49ba21e89467 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -2113,7 +2113,7 @@ static void nvme_update_disk_info(struct gendisk *disk,
nvme_config_discard(disk, ns);
nvme_config_write_zeroes(disk, ns);
- if (id->nsattr & NVME_NS_ATTR_RO)
+ if (id->nsattr & NVME_NS_ATTR_RO || test_bit(NVME_NS_FORCE_RO, &ns->flags))
set_disk_ro(disk, true);
}
diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h
index ae017f727798..bfcedfa4b057 100644
--- a/drivers/nvme/host/nvme.h
+++ b/drivers/nvme/host/nvme.h
@@ -452,6 +452,7 @@ struct nvme_ns {
#define NVME_NS_REMOVING 0
#define NVME_NS_DEAD 1
#define NVME_NS_ANA_PENDING 2
+#define NVME_NS_FORCE_RO 3
struct nvme_fault_inject fault_inject;
diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c
index 67e87e9f306f..c2b557452483 100644
--- a/drivers/nvme/host/zns.c
+++ b/drivers/nvme/host/zns.c
@@ -55,12 +55,16 @@ int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf)
int status;
/* Driver requires zone append support */
- if (!(le32_to_cpu(log->iocs[nvme_cmd_zone_append]) &
- NVME_CMD_EFFECTS_CSUPP)) {
+ if ((le32_to_cpu(log->iocs[nvme_cmd_zone_append]) & NVME_CMD_EFFECTS_CSUPP)) {
+ if (test_and_clear_bit(NVME_NS_FORCE_RO, &ns->flags))
+ dev_warn(ns->ctrl->device,
+ "append supported for zoned namespace:%d. Remove read-only mode\n",
+ ns->head->ns_id);
+ } else {
+ set_bit(NVME_NS_FORCE_RO, &ns->flags);
dev_warn(ns->ctrl->device,
- "append not supported for zoned namespace:%d\n",
- ns->head->ns_id);
- return -EINVAL;
+ "append not supported for zoned namespace:%d. Forcing to read-only mode\n",
+ ns->head->ns_id);
}
/* Lazily query controller append limit for the first zoned namespace */
--
2.17.1
More information about the Linux-nvme
mailing list