| Message ID | 146135444310.4228.7151000859132505494.stgit@dwillia2-desk3.amr.corp.intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 4/22/2016 3:47 PM, Dan Williams wrote: > In the case when a filesystem is mounted on a namespace targeted for > reconfiguration or destruction, emit a message so the user knows to > unmount the filesystem. > > Signed-off-by: Dan Williams <dan.j.williams@intel.com> > --- > builtin-xaction-namespace.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/builtin-xaction-namespace.c b/builtin-xaction-namespace.c > index 53ff82bc2c8d..e2e6d518c64d 100644 > --- a/builtin-xaction-namespace.c > +++ b/builtin-xaction-namespace.c > @@ -550,6 +550,8 @@ static int namespace_destroy(struct ndctl_region *region, > * stopping the !bdev case from racing to mount an fs or > * re-enabling the namepace. > */ > + error("%s: %s is mounted, failing operation\n", > + devname, bdev); I'm seeing this error message in cases where the device is not mounted, such as when I forget to run ndctl as root/sudo. That's a more likely reason that the open() may fail. Perhaps you could check the errno from the open() failure and display an error based on that? You might simply do a perror(). -- ljk > return -EBUSY; > } > } > > _______________________________________________ > Linux-nvdimm mailing list > Linux-nvdimm@lists.01.org > https://lists.01.org/mailman/listinfo/linux-nvdimm >
On Fri, Apr 22, 2016 at 4:00 PM, Linda Knippers <linda.knippers@hpe.com> wrote: > On 4/22/2016 3:47 PM, Dan Williams wrote: >> In the case when a filesystem is mounted on a namespace targeted for >> reconfiguration or destruction, emit a message so the user knows to >> unmount the filesystem. >> >> Signed-off-by: Dan Williams <dan.j.williams@intel.com> >> --- >> builtin-xaction-namespace.c | 2 ++ >> 1 file changed, 2 insertions(+) >> >> diff --git a/builtin-xaction-namespace.c b/builtin-xaction-namespace.c >> index 53ff82bc2c8d..e2e6d518c64d 100644 >> --- a/builtin-xaction-namespace.c >> +++ b/builtin-xaction-namespace.c >> @@ -550,6 +550,8 @@ static int namespace_destroy(struct ndctl_region *region, >> * stopping the !bdev case from racing to mount an fs or >> * re-enabling the namepace. >> */ >> + error("%s: %s is mounted, failing operation\n", >> + devname, bdev); > > I'm seeing this error message in cases where the device is not mounted, > such as when I forget to run ndctl as root/sudo. That's a more > likely reason that the open() may fail. Perhaps you could check the > errno from the open() failure and display an error based on that? > You might simply do a perror(). > Ah, thanks Linda! Yes, checking for EACESS vs EBUSY should do the trick here.
diff --git a/builtin-xaction-namespace.c b/builtin-xaction-namespace.c index 53ff82bc2c8d..e2e6d518c64d 100644 --- a/builtin-xaction-namespace.c +++ b/builtin-xaction-namespace.c @@ -550,6 +550,8 @@ static int namespace_destroy(struct ndctl_region *region, * stopping the !bdev case from racing to mount an fs or * re-enabling the namepace. */ + error("%s: %s is mounted, failing operation\n", + devname, bdev); return -EBUSY; } }
In the case when a filesystem is mounted on a namespace targeted for reconfiguration or destruction, emit a message so the user knows to unmount the filesystem. Signed-off-by: Dan Williams <dan.j.williams@intel.com> --- builtin-xaction-namespace.c | 2 ++ 1 file changed, 2 insertions(+)