| Message ID | 20220609211130.5108-7-logang@deltatee.com (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
| Series | Bug fixes and testing improvments | expand |
On Thu, 9 Jun 2022 15:11:22 -0600 Logan Gunthorpe <logang@deltatee.com> wrote: > The commit noted below globally adds a parameter to the -r option but missed > the fact that -r is used for another purpose: --remove. > > After that commit, a command such as: > > mdadm /dev/md0 -r /dev/loop0 > > will do nothing seeing the device parameter will be consumed as a > argument to the -r option; thus, there will only be one device > seen one the command line, devs_found will only be 1 and nothing will > happen. > > This caused the 01r5integ and 01raid6integ tests to hang indefinitely > as mdadm did not remove the failed device. With the device not removed, > it would not be readded. Then the loop waiting for the array status to > change would loop forever. > > To fix this, revert the changes in the noted patch and create a new subopt > type for the monitor mode with parameters required for -r. > > Fixes: 546047688e1c ("mdadm: fix coredump of mdadm --monitor -r") > Cc: Wu Guanghao <wuguanghao3@huawei.com> > Cc: Mariusz Tkaczyk <mariusz.tkaczyk@linux.intel.com> > Signed-off-by: Logan Gunthorpe <logang@deltatee.com> > --- > ReadMe.c | 7 ++++--- > mdadm.c | 1 + > mdadm.h | 1 + > 3 files changed, 6 insertions(+), 3 deletions(-) > > diff --git a/ReadMe.c b/ReadMe.c > index 8f873c4895da..556104f75d72 100644 > --- a/ReadMe.c > +++ b/ReadMe.c > @@ -81,11 +81,12 @@ char Version[] = "mdadm - v" VERSION " - " VERS_DATE > EXTRAVERSION "\n"; > * found, it is started. > */ > > -char > short_options[]="-ABCDEFGIQhVXYWZ:vqbc:i:l:p:m:r:n:x:u:c:d:z:U:N:safRSow1tye:k"; > +char > short_options[]="-ABCDEFGIQhVXYWZ:vqbc:i:l:p:m:n:x:u:c:d:z:U:N:sarfRSow1tye:k"; > +char > short_monitor_options[]="-ABCDEFGIQhVXYWZ:vqbc:i:l:p:m:r:n:x:u:c:d:z:U:N:safRSow1tye:k"; > char short_bitmap_options[]= > - > "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:r:n:x:u:c:d:z:U:N:sarfRSow1tye:k:"; > + > "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:n:x:u:c:d:z:U:N:sarfRSow1tye:k:"; char > short_bitmap_auto_options[]= > - > "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:r:n:x:u:c:d:z:U:N:sa:rfRSow1tye:k:"; > + > "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:n:x:u:c:d:z:U:N:sa:rfRSow1tye:k:"; > struct option long_options[] = { > {"manage", 0, 0, ManageOpt}, > diff --git a/mdadm.c b/mdadm.c > index be40686cf91b..d0c5e6def901 100644 > --- a/mdadm.c > +++ b/mdadm.c > @@ -227,6 +227,7 @@ int main(int argc, char *argv[]) > shortopt = short_bitmap_auto_options; > break; > case 'F': newmode = MONITOR; > + shortopt = short_monitor_options; > break; > case 'G': newmode = GROW; > shortopt = short_bitmap_options; > diff --git a/mdadm.h b/mdadm.h > index 09915a0009d9..559da3f6f440 100644 > --- a/mdadm.h > +++ b/mdadm.h > @@ -419,6 +419,7 @@ enum mode { > }; > > extern char short_options[]; > +extern char short_monitor_options[]; > extern char short_bitmap_options[]; > extern char short_bitmap_auto_options[]; > extern struct option long_options[]; Jes applied Nigel's revert but I cannot see it on repository yet. I consider adding short monitor options as valuable. Could you apply revert manually and then adopt your patch? https://lore.kernel.org/linux-raid/5f9a4417-d044-a87e-3945-2c6b29278d8c@trained-monkey.org/#t Thanks, Mariusz
Dear Logan, Thank you for the patch. There is a small typo in *regression* in the commit message summary. Kind regards, Paul
diff --git a/ReadMe.c b/ReadMe.c index 8f873c4895da..556104f75d72 100644 --- a/ReadMe.c +++ b/ReadMe.c @@ -81,11 +81,12 @@ char Version[] = "mdadm - v" VERSION " - " VERS_DATE EXTRAVERSION "\n"; * found, it is started. */ -char short_options[]="-ABCDEFGIQhVXYWZ:vqbc:i:l:p:m:r:n:x:u:c:d:z:U:N:safRSow1tye:k"; +char short_options[]="-ABCDEFGIQhVXYWZ:vqbc:i:l:p:m:n:x:u:c:d:z:U:N:sarfRSow1tye:k"; +char short_monitor_options[]="-ABCDEFGIQhVXYWZ:vqbc:i:l:p:m:r:n:x:u:c:d:z:U:N:safRSow1tye:k"; char short_bitmap_options[]= - "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:r:n:x:u:c:d:z:U:N:sarfRSow1tye:k:"; + "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:n:x:u:c:d:z:U:N:sarfRSow1tye:k:"; char short_bitmap_auto_options[]= - "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:r:n:x:u:c:d:z:U:N:sa:rfRSow1tye:k:"; + "-ABCDEFGIQhVXYWZ:vqb:c:i:l:p:m:n:x:u:c:d:z:U:N:sa:rfRSow1tye:k:"; struct option long_options[] = { {"manage", 0, 0, ManageOpt}, diff --git a/mdadm.c b/mdadm.c index be40686cf91b..d0c5e6def901 100644 --- a/mdadm.c +++ b/mdadm.c @@ -227,6 +227,7 @@ int main(int argc, char *argv[]) shortopt = short_bitmap_auto_options; break; case 'F': newmode = MONITOR; + shortopt = short_monitor_options; break; case 'G': newmode = GROW; shortopt = short_bitmap_options; diff --git a/mdadm.h b/mdadm.h index 09915a0009d9..559da3f6f440 100644 --- a/mdadm.h +++ b/mdadm.h @@ -419,6 +419,7 @@ enum mode { }; extern char short_options[]; +extern char short_monitor_options[]; extern char short_bitmap_options[]; extern char short_bitmap_auto_options[]; extern struct option long_options[];
The commit noted below globally adds a parameter to the -r option but missed the fact that -r is used for another purpose: --remove. After that commit, a command such as: mdadm /dev/md0 -r /dev/loop0 will do nothing seeing the device parameter will be consumed as a argument to the -r option; thus, there will only be one device seen one the command line, devs_found will only be 1 and nothing will happen. This caused the 01r5integ and 01raid6integ tests to hang indefinitely as mdadm did not remove the failed device. With the device not removed, it would not be readded. Then the loop waiting for the array status to change would loop forever. To fix this, revert the changes in the noted patch and create a new subopt type for the monitor mode with parameters required for -r. Fixes: 546047688e1c ("mdadm: fix coredump of mdadm --monitor -r") Cc: Wu Guanghao <wuguanghao3@huawei.com> Cc: Mariusz Tkaczyk <mariusz.tkaczyk@linux.intel.com> Signed-off-by: Logan Gunthorpe <logang@deltatee.com> --- ReadMe.c | 7 ++++--- mdadm.c | 1 + mdadm.h | 1 + 3 files changed, 6 insertions(+), 3 deletions(-)