* [PATCH] nvme-cli/format: use existing LBAF if none requested
@ 2017-03-14 15:54 Keith Busch
2017-03-14 16:23 ` Christoph Hellwig
0 siblings, 1 reply; 2+ messages in thread
From: Keith Busch @ 2017-03-14 15:54 UTC (permalink / raw)
We previously default to LBAF 0 if the user didn't specify one. If a
specific LBAF is not explicitly requested, most people expect format
to not change their currently in use LBAF. This patch uses the current
format if we have a namespace to query. If we don't have a namespace,
and the user did not request a specific LBAF, the value will default to
0 as before.
https://github.com/linux-nvme/nvme-cli/issues/164
Signed-off-by: Keith Busch <keith.busch at intel.com>
---
nvme.c | 24 +++++++++++++++++++++---
1 file changed, 21 insertions(+), 3 deletions(-)
diff --git a/nvme.c b/nvme.c
index 17318b4..c56ebb4 100644
--- a/nvme.c
+++ b/nvme.c
@@ -1397,7 +1397,9 @@ static int format(int argc, char **argv, struct command *cmd, struct plugin *plu
const char *ms = "[0-1]: extended format off/on";
const char *reset = "Automatically reset the controller after successful format";
const char *timeout = "timeout value, in milliseconds";
+ struct nvme_id_ns ns;
int err, fd;
+ __u8 prev_lbaf = 0;
struct config {
__u32 namespace_id;
@@ -1413,7 +1415,7 @@ static int format(int argc, char **argv, struct command *cmd, struct plugin *plu
struct config cfg = {
.namespace_id = 0xffffffff,
.timeout = 120000,
- .lbaf = 0,
+ .lbaf = 0xff,
.ses = 0,
.pi = 0,
.reset = 0,
@@ -1435,6 +1437,24 @@ static int format(int argc, char **argv, struct command *cmd, struct plugin *plu
if (fd < 0)
return fd;
+ if (S_ISBLK(nvme_stat.st_mode))
+ cfg.namespace_id = get_nsid(fd);
+ if (cfg.namespace_id) {
+ err = nvme_identify_ns(fd, cfg.namespace_id, 0, &ns);
+ if (err) {
+ if (err < 0)
+ perror("identify-namespace");
+ else
+ fprintf(stderr,
+ "NVME Admin command error:%s(%x)\n",
+ nvme_status_to_string(err), err);
+ return err;
+ }
+ prev_lbaf = ns.flbas & 0xf;
+ }
+ if (cfg.lbaf == 0xff)
+ cfg.lbaf = prev_lbaf;
+
/* ses & pi checks set to 7 for forward-compatibility */
if (cfg.ses > 7) {
fprintf(stderr, "invalid secure erase settings:%d\n", cfg.ses);
@@ -1456,8 +1476,6 @@ static int format(int argc, char **argv, struct command *cmd, struct plugin *plu
fprintf(stderr, "invalid ms:%d\n", cfg.ms);
return EINVAL;
}
- if (S_ISBLK(nvme_stat.st_mode))
- cfg.namespace_id = get_nsid(fd);
err = nvme_format(fd, cfg.namespace_id, cfg.lbaf, cfg.ses, cfg.pi,
cfg.pil, cfg.ms, cfg.timeout);
--
2.5.5
^ permalink raw reply related [flat|nested] 2+ messages in thread
* [PATCH] nvme-cli/format: use existing LBAF if none requested
2017-03-14 15:54 [PATCH] nvme-cli/format: use existing LBAF if none requested Keith Busch
@ 2017-03-14 16:23 ` Christoph Hellwig
0 siblings, 0 replies; 2+ messages in thread
From: Christoph Hellwig @ 2017-03-14 16:23 UTC (permalink / raw)
This looks fine to me:
Reviewed-by: Christoph Hellwig <hch at lst.de>
Btw, we really should do something about the global format behavior
bits in FNA - if bit 0 or 1 are set we should fail formats that
are specific to a namespace and not global with a warnings.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2017-03-14 16:23 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-03-14 15:54 [PATCH] nvme-cli/format: use existing LBAF if none requested Keith Busch
2017-03-14 16:23 ` Christoph Hellwig
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).