| Message ID | 20210708100104.168348-1-aneesh.kumar@linux.ibm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | ndctl: Avoid confusing error message when operating on all the namespaces | expand |
Jeff had posted one patch to fix similar issue https://lore.kernel.org/linux-nvdimm/x49r1lohpty.fsf@segfault.boston.devel.redhat.com/T/#u Hi Dan/Visha Could we make some progress on this issue? On Thu, Jul 8, 2021 at 6:41 PM Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> wrote: > > With only seed namespace present, ndctl results in confusing error messages as > below. > > ndctl# ./ndctl/ndctl enable-namespace all > error enabling namespaces: No such device or address > enabled 0 namespaces > > ndctl# ./ndctl/ndctl disable-namespace all > disabled 3 namespaces > > ndctl# ./ndctl/ndctl destroy-namespace all -f > Error: destroy namespace: namespace1.0 failed to enable for zeroing, continuing > > Error: destroy namespace: namespace1.1 failed to enable for zeroing, continuing > > Error: destroy namespace: namespace0.0 failed to enable for zeroing, continuing > > destroyed 0 namespaces > ndctl# > > With the patch we get > ndctl# ./ndctl/ndctl disable-namespace all > disabled 0 namespaces > > ndctl# ./ndctl/ndctl enable-namespace all > enabled 0 namespaces > > ndctl# ./ndctl/ndctl destroy-namespace all -f > destroyed 0 namespaces > ndctl# > > Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> > --- > ndctl/namespace.c | 9 ++++++++- > 1 file changed, 8 insertions(+), 1 deletion(-) > > diff --git a/ndctl/namespace.c b/ndctl/namespace.c > index 0c8df9fa8b47..c52daeae562a 100644 > --- a/ndctl/namespace.c > +++ b/ndctl/namespace.c > @@ -2205,8 +2205,15 @@ static int do_xaction_namespace(const char *namespace, > return rc; > } > ndctl_namespace_foreach_safe(region, ndns, _n) { > - ndns_name = ndctl_namespace_get_devname(ndns); > > + if (!strcmp(namespace, "all") > + && !ndctl_namespace_get_size(ndns)) { > + if (!*processed && rc) > + rc = 0; > + continue; > + } > + > + ndns_name = ndctl_namespace_get_devname(ndns); > if (strcmp(namespace, "all") != 0 > && strcmp(namespace, ndns_name) != 0) > continue; > -- > 2.31.1 > >
On 7/13/21 5:49 AM, Yi Zhang wrote: > Jeff had posted one patch to fix similar issue > https://lore.kernel.org/linux-nvdimm/x49r1lohpty.fsf@segfault.boston.devel.redhat.com/T/#u > I missed that patch. I am wondering whether that would result in ndctl not reporting error when we use the seed namespace name directly with destroy-namespace command > Hi Dan/Visha > Could we make some progress on this issue? > > > On Thu, Jul 8, 2021 at 6:41 PM Aneesh Kumar K.V > <aneesh.kumar@linux.ibm.com> wrote: >> >> With only seed namespace present, ndctl results in confusing error messages as >> below. >> >> ndctl# ./ndctl/ndctl enable-namespace all >> error enabling namespaces: No such device or address >> enabled 0 namespaces >> >> ndctl# ./ndctl/ndctl disable-namespace all >> disabled 3 namespaces >> >> ndctl# ./ndctl/ndctl destroy-namespace all -f >> Error: destroy namespace: namespace1.0 failed to enable for zeroing, continuing >> >> Error: destroy namespace: namespace1.1 failed to enable for zeroing, continuing >> >> Error: destroy namespace: namespace0.0 failed to enable for zeroing, continuing >> >> destroyed 0 namespaces >> ndctl# >> >> With the patch we get >> ndctl# ./ndctl/ndctl disable-namespace all >> disabled 0 namespaces >> >> ndctl# ./ndctl/ndctl enable-namespace all >> enabled 0 namespaces >> >> ndctl# ./ndctl/ndctl destroy-namespace all -f >> destroyed 0 namespaces >> ndctl# >> >> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> >> --- >> ndctl/namespace.c | 9 ++++++++- >> 1 file changed, 8 insertions(+), 1 deletion(-) >> >> diff --git a/ndctl/namespace.c b/ndctl/namespace.c >> index 0c8df9fa8b47..c52daeae562a 100644 >> --- a/ndctl/namespace.c >> +++ b/ndctl/namespace.c >> @@ -2205,8 +2205,15 @@ static int do_xaction_namespace(const char *namespace, >> return rc; >> } >> ndctl_namespace_foreach_safe(region, ndns, _n) { >> - ndns_name = ndctl_namespace_get_devname(ndns); >> >> + if (!strcmp(namespace, "all") >> + && !ndctl_namespace_get_size(ndns)) { >> + if (!*processed && rc) >> + rc = 0; >> + continue; >> + } >> + >> + ndns_name = ndctl_namespace_get_devname(ndns); >> if (strcmp(namespace, "all") != 0 >> && strcmp(namespace, ndns_name) != 0) >> continue; >> -- >> 2.31.1 >> >> > >
On Mon, Jul 12, 2021 at 5:20 PM Yi Zhang <yi.zhang@redhat.com> wrote: > > Jeff had posted one patch to fix similar issue > https://lore.kernel.org/linux-nvdimm/x49r1lohpty.fsf@segfault.boston.devel.redhat.com/T/#u > > Hi Dan/Visha > Could we make some progress on this issue? Apologies, we had some internal administrivia to address, but are getting back to regular releases now and catching up on the backlog.
On Tue, Jul 13, 2021 at 1:16 PM Dan Williams <dan.j.williams@intel.com> wrote: > > On Mon, Jul 12, 2021 at 5:20 PM Yi Zhang <yi.zhang@redhat.com> wrote: > > > > Jeff had posted one patch to fix similar issue > > https://lore.kernel.org/linux-nvdimm/x49r1lohpty.fsf@segfault.boston.devel.redhat.com/T/#u > > > > Hi Dan/Visha > > Could we make some progress on this issue? > > Apologies, we had some internal administrivia to address, but are > getting back to regular releases now and catching up on the backlog. > No worries, thanks for the update :)
Dan Williams <dan.j.williams@intel.com> writes: > On Mon, Jul 12, 2021 at 5:20 PM Yi Zhang <yi.zhang@redhat.com> wrote: >> >> Jeff had posted one patch to fix similar issue >> https://lore.kernel.org/linux-nvdimm/x49r1lohpty.fsf@segfault.boston.devel.redhat.com/T/#u >> >> Hi Dan/Visha >> Could we make some progress on this issue? > > Apologies, we had some internal administrivia to address, but are > getting back to regular releases now and catching up on the backlog. Any update on this patch? Without this I get failures with https://github.com/avocado-framework-tests/avocado-misc-tests/blob/master/memory/ndctl.py because the return status is 1 for disable-namespace command with only seed namespace. -aneesh
diff --git a/ndctl/namespace.c b/ndctl/namespace.c index 0c8df9fa8b47..c52daeae562a 100644 --- a/ndctl/namespace.c +++ b/ndctl/namespace.c @@ -2205,8 +2205,15 @@ static int do_xaction_namespace(const char *namespace, return rc; } ndctl_namespace_foreach_safe(region, ndns, _n) { - ndns_name = ndctl_namespace_get_devname(ndns); + if (!strcmp(namespace, "all") + && !ndctl_namespace_get_size(ndns)) { + if (!*processed && rc) + rc = 0; + continue; + } + + ndns_name = ndctl_namespace_get_devname(ndns); if (strcmp(namespace, "all") != 0 && strcmp(namespace, ndns_name) != 0) continue;
With only seed namespace present, ndctl results in confusing error messages as below. ndctl# ./ndctl/ndctl enable-namespace all error enabling namespaces: No such device or address enabled 0 namespaces ndctl# ./ndctl/ndctl disable-namespace all disabled 3 namespaces ndctl# ./ndctl/ndctl destroy-namespace all -f Error: destroy namespace: namespace1.0 failed to enable for zeroing, continuing Error: destroy namespace: namespace1.1 failed to enable for zeroing, continuing Error: destroy namespace: namespace0.0 failed to enable for zeroing, continuing destroyed 0 namespaces ndctl# With the patch we get ndctl# ./ndctl/ndctl disable-namespace all disabled 0 namespaces ndctl# ./ndctl/ndctl enable-namespace all enabled 0 namespaces ndctl# ./ndctl/ndctl destroy-namespace all -f destroyed 0 namespaces ndctl# Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> --- ndctl/namespace.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)