From mboxrd@z Thu Jan  1 00:00:00 1970
From: Luca Berra <bluca@comedia.it>
Subject: Re: [PATCH 1/2] imsm: delete subarray functionality
Date: Thu, 1 Apr 2010 08:09:22 +0200
Message-ID: <20100401060922.GA15461@maude.comedia.it>
References: <66C59AD0932712458090B447266D638CD3C7EAF4@irsmsx504.ger.corp.intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Return-path: <linux-raid-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <66C59AD0932712458090B447266D638CD3C7EAF4@irsmsx504.ger.corp.intel.com>
Sender: linux-raid-owner@vger.kernel.org
To: "Hawrylewicz Czarnowski, Przemyslaw" <przemyslaw.hawrylewicz.czarnowski@intel.com>
Cc: Neil Brown <neilb@suse.de>, "linux-raid@vger.kernel.org" <linux-raid@vger.kernel.org>, "Williams, Dan J" <dan.j.williams@intel.com>, "Ciechanowski, Ed" <ed.ciechanowski@intel.com>
List-Id: linux-raid.ids

On Wed, Mar 31, 2010 at 09:50:40PM +0100, Hawrylewicz Czarnowski, Przemyslaw wrote:
>Patch also adds some utilities to simple sysfs manipulation and
>to allow locking of map file during updates.
a 38k patch is very difficult to review, i would have split it. the
utilities would have been a candidate for a different patch
also there are some changes i would define either bug-fixes or cleanup
they should be in a different patch.
you forgot to update the man pages.


>diff --git a/Kill.c b/Kill.c
>index e738978..e5d5dcb 100644
>--- a/Kill.c
>+++ b/Kill.c
>@@ -79,3 +79,61 @@ int Kill(char *dev, struct supertype *st, int force, int quiet, int noexcl)
>        close(fd);
>        return rv;
> }
>+
>+
>+int Kill_Subarray(mddev_dev_t devlist, int force, int quiet, struct mddev_ident_s *ident)
>+{
>+       int     rv;
>+       struct array_dev_list *dl = NULL;
>+       int i, found;
>+       struct mdinfo info;
>+
>+       if ((dl = check_devices(devlist, force, quiet, "imsm")) == NULL) {
>+               return 1;
>+       }
i dont know if other container types will ever support removing (or
renaming) subarrays. but i am not sure that forcing imsm here versus
checking the existance of ss->delete_subarray function is better.
....

>diff --git a/mapfile.c b/mapfile.c
>index 366ebe3..fe885c5 100644
>--- a/mapfile.c
>+++ b/mapfile.c
the above is a candidate for a separate patch
....

>diff --git a/mdstat.c b/mdstat.c
>index 4a9f370..bb6179d 100644
>--- a/mdstat.c
>+++ b/mdstat.c
ditto
....

>diff --git a/sysfs.c b/sysfs.c
>index ebf9d8a..a2353d9 100644
>--- a/sysfs.c
>+++ b/sysfs.c
ditto
....

>diff --git a/util.c b/util.c
>index d292a66..f1d30f3 100644
>--- a/util.c
>+++ b/util.c
as above, but also:

>+char *dev2devname(int maj, int min, char *buf)
what is this for, completeness? it is unused in your code
...

> int mdmon_pid(int devnum)
>@@ -1502,7 +1523,11 @@ int mdmon_pid(int devnum)
>        char pid[10];
>        int fd;
>        int n;
>-       sprintf(path, "%s/%s.pid", pid_dir, devnum2devname(devnum));
>+       char *devname = devnum2devname(devnum);
>+
>+       sprintf(path, "%s/%s.pid", pid_dir, devname);
>+       free(devname);
>+
>        fd = open(path, O_RDONLY | O_NOATIME, 0);
>
>        if (fd < 0)
why?

>+int validate_super(struct supertype *st, char *name, int quiet, char *meta_type)
>+{
>+       if (st == NULL) {
>+               if (!quiet && name)
>+                       fprintf(stderr, Name ": Unrecognised md component device - %s\n", name);
>+               return 1;
>+       }
>+
>+       if (meta_type && strcmp(st->ss->name, meta_type)) {
>+               if (!quiet)
>+                       fprintf(stderr, Name ": Subarray delete not supported for '%s' superblock\n",
>+                               st->ss->name);
i believe this error message belongs to the caller actually even two
levels above, not to the utility function.

....

>+struct array_dev_list *check_devices(mddev_dev_t devlist, int force, int quiet, char *meta_type)
<snip>

>+               if (validate_super(st, dv->devname, quiet, "imsm")) {
i believe it should read:
                if (validate_super(st, dv->devname, quiet, meta_type)) {


regards,
L.
-- 
Luca Berra -- bluca@comedia.it
           Communication Media & Services S.r.l.
    /"\
    \ /     ASCII RIBBON CAMPAIGN
     X        AGAINST HTML MAIL
    / \