From: mwilck@arcor.de
To: neilb@suse.de, linux-raid@vger.kernel.org
Cc: mwilck@arcor.de
Subject: [PATCH 08/27] DDF: ddf_set_disk: move status logic to separate function
Date: Wed, 3 Jul 2013 22:27:48 +0200 [thread overview]
Message-ID: <1372883287-8859-9-git-send-email-mwilck@arcor.de> (raw)
In-Reply-To: <1372883287-8859-1-git-send-email-mwilck@arcor.de>
Moved code to determine RAID status to a separate function
get_bvd_status(). I need this to account for secondary RAID level.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
---
super-ddf.c | 94 +++++++++++++++++++++++++++++++++-------------------------
1 files changed, 53 insertions(+), 41 deletions(-)
diff --git a/super-ddf.c b/super-ddf.c
index c448bff..322ad5e 100644
--- a/super-ddf.c
+++ b/super-ddf.c
@@ -1515,7 +1515,7 @@ bad:
}
#endif
-static int find_phys(struct ddf_super *ddf, __u32 phys_refnum)
+static int find_phys(const struct ddf_super *ddf, __u32 phys_refnum)
{
/* Find the entry in phys_disk which has the given refnum
* and return it's index
@@ -3812,9 +3812,55 @@ static int ddf_set_array_state(struct active_array *a, int consistent)
return consistent;
}
-#define container_of(ptr, type, member) ({ \
- const typeof( ((type *)0)->member ) *__mptr = (ptr); \
- (type *)( (char *)__mptr - offsetof(type,member) );})
+static int get_bvd_state(const struct ddf_super *ddf,
+ const struct vd_config *vc)
+{
+ unsigned int i, n_bvd, working = 0;
+ unsigned int n_prim = __be16_to_cpu(vc->prim_elmnt_count);
+ int pd, st, state;
+ for (i = 0; i < n_prim; i++) {
+ if (!find_index_in_bvd(ddf, vc, i, &n_bvd))
+ continue;
+ pd = find_phys(ddf, vc->phys_refnum[n_bvd]);
+ if (pd < 0)
+ continue;
+ st = __be16_to_cpu(ddf->phys->entries[pd].state);
+ if ((st & (DDF_Online|DDF_Failed|DDF_Rebuilding))
+ == DDF_Online)
+ working++;
+ }
+
+ state = DDF_state_degraded;
+ if (working == n_prim)
+ state = DDF_state_optimal;
+ else
+ switch (vc->prl) {
+ case DDF_RAID0:
+ case DDF_CONCAT:
+ case DDF_JBOD:
+ state = DDF_state_failed;
+ break;
+ case DDF_RAID1:
+ if (working == 0)
+ state = DDF_state_failed;
+ else if (working >= 2)
+ state = DDF_state_part_optimal;
+ break;
+ case DDF_RAID4:
+ case DDF_RAID5:
+ if (working < n_prim - 1)
+ state = DDF_state_failed;
+ break;
+ case DDF_RAID6:
+ if (working < n_prim - 2)
+ state = DDF_state_failed;
+ else if (working == n_prim - 1)
+ state = DDF_state_part_optimal;
+ break;
+ }
+ return state;
+}
+
/*
* The state of each disk is stored in the global phys_disk structure
* in phys_disk.entries[n].state.
@@ -3837,7 +3883,6 @@ static void ddf_set_disk(struct active_array *a, int n, int state)
struct vd_config *vc = find_vdcr(ddf, inst, (unsigned int)n,
&n_bvd, &vcl);
int pd;
- int i, st, working;
struct mdinfo *mdi;
struct dl *dl;
@@ -3901,43 +3946,10 @@ static void ddf_set_disk(struct active_array *a, int n, int state)
* It needs to be one of "optimal", "degraded", "failed".
* I don't understand 'deleted' or 'missing'.
*/
- working = 0;
- for (i=0; i < a->info.array.raid_disks; i++) {
- pd = find_phys(ddf, vc->phys_refnum[i]);
- if (pd < 0)
- continue;
- st = __be16_to_cpu(ddf->phys->entries[pd].state);
- if ((st & (DDF_Online|DDF_Failed|DDF_Rebuilding))
- == DDF_Online)
- working++;
+ state = get_bvd_state(ddf, vc);
+ if (vc->sec_elmnt_count > 1) {
+ /* treat secondary level */
}
- state = DDF_state_degraded;
- if (working == a->info.array.raid_disks)
- state = DDF_state_optimal;
- else switch(vc->prl) {
- case DDF_RAID0:
- case DDF_CONCAT:
- case DDF_JBOD:
- state = DDF_state_failed;
- break;
- case DDF_RAID1:
- if (working == 0)
- state = DDF_state_failed;
- else if (working == 2 && state == DDF_state_degraded)
- state = DDF_state_part_optimal;
- break;
- case DDF_RAID4:
- case DDF_RAID5:
- if (working < a->info.array.raid_disks-1)
- state = DDF_state_failed;
- break;
- case DDF_RAID6:
- if (working < a->info.array.raid_disks-2)
- state = DDF_state_failed;
- else if (working == a->info.array.raid_disks-1)
- state = DDF_state_part_optimal;
- break;
- }
if (ddf->virt->entries[inst].state !=
((ddf->virt->entries[inst].state & ~DDF_state_mask)
--
1.7.1
next prev parent reply other threads:[~2013-07-03 20:27 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-07-03 20:27 DDF/RAID10 patch series mwilck
2013-07-03 20:27 ` [PATCH 01/27] DDF (cleanup): use a common macro for failed searches mwilck
2013-07-03 20:27 ` [PATCH 02/27] DDF: check_secondary: fix treatment of missing BVDs mwilck
2013-07-03 20:27 ` [PATCH 03/27] DDF: load_ddf_headers: use secondary header as fallback mwilck
2013-07-08 5:43 ` NeilBrown
2013-07-08 20:56 ` Martin Wilck
2013-07-03 20:27 ` [PATCH 04/27] DDF: handle "open flag" according to spec mwilck
2013-07-03 20:27 ` [PATCH 05/27] DDF: Implement store_super_ddf mwilck
2013-07-08 5:48 ` NeilBrown
2013-07-10 17:37 ` Martin Wilck
2013-07-15 5:31 ` NeilBrown
2013-07-18 18:48 ` [PATCH 1/3] DDF: increase seq number in ddf_set_updates_pending mwilck
2013-07-18 18:49 ` [PATCH 2/3] DDF: make "null_aligned" a static buffer mwilck
2013-07-18 18:49 ` [PATCH 3/3] DDF: factor out writing super block to single disk mwilck
2013-07-18 18:51 ` [PATCH 05/27] DDF: Implement store_super_ddf Martin Wilck
2013-07-03 20:27 ` [PATCH 06/27] DDF: ddf_open_new: implement minimal consistency check mwilck
2013-07-03 20:27 ` [PATCH 07/27] DDF: find_vdcr: account for secondary RAID level mwilck
2013-07-08 6:16 ` NeilBrown
2013-07-08 21:03 ` Martin Wilck
2013-07-03 20:27 ` mwilck [this message]
2013-07-03 20:27 ` [PATCH 09/27] DDF: get_svd_state: Status logic " mwilck
2013-07-03 20:27 ` [PATCH 10/27] DDF: allow empty slots in virt disk table mwilck
2013-07-03 20:27 ` [PATCH 11/27] DDF: layout_ddf2md: new DDF->md RAID layout conversion mwilck
2013-07-08 6:48 ` NeilBrown
2013-07-08 21:06 ` Martin Wilck
2013-07-03 20:27 ` [PATCH 12/27] DDF: layout_md2ddf: new md->DDF " mwilck
2013-07-08 6:28 ` NeilBrown
2013-07-03 20:27 ` [PATCH 13/27] DDF: Simplify allocation of "other BVDs" mwilck
2013-07-03 20:27 ` [PATCH 14/27] DDF: init_super_ddf_bvd: initialize other bvds mwilck
2013-07-03 20:27 ` [PATCH 15/27] DDF: validate_geometry_ddf: support RAID10 mwilck
2013-07-03 20:27 ` [PATCH 16/27] DDF: use LBA_OFFSET macro instead of lba_offset field mwilck
2013-07-03 20:27 ` [PATCH 17/27] DDF: get_extents: support secondary RAID level mwilck
2013-07-03 20:27 ` [PATCH 18/27] DDF: add_to_super_ddf: allow empty slots in phys disk table mwilck
2013-07-03 20:27 ` [PATCH 19/27] DDF: add_to_super_ddf: Use same amount of workspace as other disks mwilck
2013-07-03 20:28 ` [PATCH 20/27] DDF: add_to_super_ddf: RAID10 changes mwilck
2013-07-03 20:28 ` [PATCH 21/27] DDF: add_to_super_ddf_bvd: use get_svd_state() mwilck
2013-07-03 20:28 ` [PATCH 22/27] DDF: getinfo_super_ddf_bvd: lba_offset calculation for RAID10 mwilck
2013-07-03 20:28 ` [PATCH 23/27] DDF: guid_str: convenience function to print GUID for debugging mwilck
2013-07-03 20:28 ` [PATCH 24/27] DDF: ddf_set_array_state: more meaningful output mwilck
2013-07-03 20:28 ` [PATCH 25/27] DDF: ddf_process_update: handle update of conf records for SVD mwilck
2013-07-03 20:28 ` [PATCH 26/27] DDF: ddf_process_update: Fix vlist treatment for SVDs mwilck
2013-07-03 20:28 ` [PATCH 27/27] tests/10ddf-create: add RAID 10 array mwilck
2013-07-04 10:10 ` DDF/RAID10 patch series NeilBrown
2013-07-08 6:53 ` NeilBrown
2013-07-08 21:50 ` Fixes for the DDF " mwilck
2013-07-08 21:50 ` [PATCH 28/39] test/10-ddf-create: fix comments mwilck
2013-07-08 21:50 ` [PATCH 29/39] Monitor: Don't write metadata in inactive array state mwilck
2013-07-08 21:50 ` [PATCH 30/39] DDF: write_init_super_ddf: don't zero superblocks for subarrays mwilck
2013-07-08 21:50 ` [PATCH 31/39] DDF: implement kill_subarray mwilck
2013-07-08 21:50 ` [PATCH 32/39] DDF: getinfo_super_ddf_bvd: identify disk by refnum mwilck
2013-07-08 21:50 ` [PATCH 33/39] DDF: getinfo_super_ddf_bvd: fix raid_disk calculation mwilck
2013-07-08 21:50 ` [PATCH 34/39] DDF: fix endianness of refnum in debug messages mwilck
2013-07-08 21:50 ` [PATCH 35/39] DDF: add debug message in add_super_ddf_bvd mwilck
2013-07-08 21:50 ` [PATCH 36/39] DDF: ddf_process_update: add debug messages fore adding VDs mwilck
2013-07-08 21:50 ` [PATCH 37/39] DDF: guid_str: more readable output mwilck
2013-07-08 21:50 ` [PATCH 38/39] DDF: ddf_process_update: some more debug messages mwilck
2013-07-08 21:50 ` [PATCH 39/39] DDF: ddf_process_update: Fix updates for SVDs mwilck
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1372883287-8859-9-git-send-email-mwilck@arcor.de \
--to=mwilck@arcor.de \
--cc=linux-raid@vger.kernel.org \
--cc=neilb@suse.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).