diff mbox series

[9/9] btrfs: Add RAID 6 recovery for a btrfs filesystem.

Message ID 20180919184040.22540-10-kreijack@libero.it (mailing list archive)
State New, archived
Headers show
Series [1/9] btrfs: Add support for reading a filesystem with a RAID 5 or RAID 6 profile. | expand

Commit Message

Goffredo Baroncelli Sept. 19, 2018, 6:40 p.m. UTC
From: Goffredo Baroncelli <kreijack@inwind.it>

Add the RAID 6 recovery, in order to use a RAID 6 filesystem even if some
disks (up to two) are missing. This code use the md RAID 6 code already
present in grub.

Signed-off-by: Goffredo Baroncelli <kreijack@inwind.it>
---
 grub-core/fs/btrfs.c | 54 +++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 48 insertions(+), 6 deletions(-)

Comments

Daniel Kiper Sept. 25, 2018, 7:20 p.m. UTC | #1
On Wed, Sep 19, 2018 at 08:40:40PM +0200, Goffredo Baroncelli wrote:
> From: Goffredo Baroncelli <kreijack@inwind.it>
>
> Add the RAID 6 recovery, in order to use a RAID 6 filesystem even if some
> disks (up to two) are missing. This code use the md RAID 6 code already
> present in grub.
>
> Signed-off-by: Goffredo Baroncelli <kreijack@inwind.it>
> ---
>  grub-core/fs/btrfs.c | 54 +++++++++++++++++++++++++++++++++++++++-----
>  1 file changed, 48 insertions(+), 6 deletions(-)
>
> diff --git a/grub-core/fs/btrfs.c b/grub-core/fs/btrfs.c
> index 55a7eeffc..400cd56b6 100644
> --- a/grub-core/fs/btrfs.c
> +++ b/grub-core/fs/btrfs.c
> @@ -30,6 +30,7 @@
>  #include <grub/i18n.h>
>  #include <grub/btrfs.h>
>  #include <grub/crypto.h>
> +#include <grub/diskfilter.h>
>
>  GRUB_MOD_LICENSE ("GPLv3+");
>
> @@ -705,11 +706,36 @@ rebuild_raid5 (char *dest, struct raid56_buffer *buffers,
>      }
>  }
>
> +static grub_err_t
> +raid6_recover_read_buffer (void *data, int disk_nr,
> +			   grub_uint64_t addr __attribute__ ((unused)),
> +			   void *dest, grub_size_t size)
> +{
> +    struct raid56_buffer *buffers = data;
> +
> +    if (!buffers[disk_nr].data_is_valid)
> +	return grub_errno = GRUB_ERR_READ_ERROR;
> +
> +    grub_memcpy(dest, buffers[disk_nr].buf, size);
> +
> +    return grub_errno = GRUB_ERR_NONE;
> +}
> +
> +static void
> +rebuild_raid6 (struct raid56_buffer *buffers, grub_uint64_t nstripes,
> +               grub_uint64_t csize, grub_uint64_t parities_pos, void *dest,
> +               grub_uint64_t stripen)
> +
> +{
> +  grub_raid6_recover_gen (buffers, nstripes, stripen, parities_pos,
> +                          dest, 0, csize, 0, raid6_recover_read_buffer);
> +}
> +
>  static grub_err_t
>  raid56_read_retry (struct grub_btrfs_data *data,
>  		   struct grub_btrfs_chunk_item *chunk,
> -		   grub_uint64_t stripe_offset,
> -		   grub_uint64_t csize, void *buf)
> +		   grub_uint64_t stripe_offset, grub_uint64_t stripen,
> +		   grub_uint64_t csize, void *buf, grub_uint64_t parities_pos)
>  {
>    struct raid56_buffer *buffers;
>    grub_uint64_t nstripes = grub_le_to_cpu16 (chunk->nstripes);
> @@ -787,6 +813,15 @@ raid56_read_retry (struct grub_btrfs_data *data,
>        ret = GRUB_ERR_READ_ERROR;
>        goto cleanup;
>      }
> +  else if (failed_devices > 2 && (chunk_type & GRUB_BTRFS_CHUNK_TYPE_RAID6))
> +    {
> +      grub_dprintf ("btrfs",
> +		    "not enough disks for raid6: total %" PRIuGRUB_UINT64_T
> +		    ", missing %" PRIuGRUB_UINT64_T "\n",
> +		    nstripes, failed_devices);
> +      ret = GRUB_ERR_READ_ERROR;
> +      goto cleanup;
> +    }
>    else
>      grub_dprintf ("btrfs",
>  		  "enough disks for RAID 5 rebuilding: total %"
> @@ -797,7 +832,7 @@ raid56_read_retry (struct grub_btrfs_data *data,
>    if (chunk_type & GRUB_BTRFS_CHUNK_TYPE_RAID5)
>      rebuild_raid5 (buf, buffers, nstripes, csize);
>    else
> -    grub_dprintf ("btrfs", "called rebuild_raid6(), NOT IMPLEMENTED\n");
> +    rebuild_raid6 (buffers, nstripes, csize, parities_pos, buf, stripen);
>
>   cleanup:
>    if (buffers)
> @@ -886,9 +921,11 @@ grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
>  	unsigned redundancy = 1;
>  	unsigned i, j;
>  	int is_raid56;
> +	grub_uint64_t parities_pos = 0;
>
> -	is_raid56 = !!(grub_le_to_cpu64 (chunk->type) &
> -		       GRUB_BTRFS_CHUNK_TYPE_RAID5);
> +        is_raid56 = !!(grub_le_to_cpu64 (chunk->type) &
> +		       (GRUB_BTRFS_CHUNK_TYPE_RAID5 |
> +		        GRUB_BTRFS_CHUNK_TYPE_RAID6));
>
>  	if (grub_le_to_cpu64 (chunk->size) <= off)
>  	  {
> @@ -1015,6 +1052,8 @@ grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
>  	       *  - stripe_offset is the disk offset,
>  	       *  - csize is the "potential" data to read. It will be reduced to
>  	       *    size if the latter is smaller.
> +	       *  - parities_pos is the position of the parity inside a row (

s/inside/in/

> +	       *    2 for P1, 3 for P2...)
>  	       */
>  	      block_nr = grub_divmod64 (off, chunk_stripe_length, &low);
>
> @@ -1030,6 +1069,9 @@ grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
>  	       */
>  	      grub_divmod64 (high + stripen, nstripes, &stripen);
>
> +	      grub_divmod64 (high + nstripes - nparities, nstripes,
> +			     &parities_pos);

I think that this math requires a bit of explanation in the comment
before grub_divmod64(). Especially I am interested in why high +
nstripes - nparities works as expected.

Daniel
Goffredo Baroncelli Sept. 26, 2018, 7:56 p.m. UTC | #2
On 25/09/2018 21.20, Daniel Kiper wrote:
> On Wed, Sep 19, 2018 at 08:40:40PM +0200, Goffredo Baroncelli wrote:
>> From: Goffredo Baroncelli <kreijack@inwind.it>
>>
[....]
>>  	       *  - stripe_offset is the disk offset,
>>  	       *  - csize is the "potential" data to read. It will be reduced to
>>  	       *    size if the latter is smaller.
>> +	       *  - parities_pos is the position of the parity inside a row (
> 
> s/inside/in/> 
>> +	       *    2 for P1, 3 for P2...)

+              *  - nparities is the number of parities (1 for RAID5, 2 for RAID6);
+              *    used only in RAID5/6 code.

>>  	       */
>>  	      block_nr = grub_divmod64 (off, chunk_stripe_length, &low);
>>
>> @@ -1030,6 +1069,9 @@ grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
>>  	       */
>>  	      grub_divmod64 (high + stripen, nstripes, &stripen);
>>
>> +	      grub_divmod64 (high + nstripes - nparities, nstripes,
>> +			     &parities_pos);
> 
> I think that this math requires a bit of explanation in the comment
> before grub_divmod64(). Especially I am interested in why high +
> nstripes - nparities works as expected.


What about

/*
 * parities_pos is equal to "(high - nparities) % nstripes" (see the diagram above).
 * However "high - nparities" might be negative (eg when high == 0) leading to an
 * incorrect computation.
 * Instead "high + nstripes - nparities" is always positive and in modulo nstripes is
 * equal to "(high - nparities) % nstripes
 */
> 
> Daniel
> 
BR
G.Baroncelli
Daniel Kiper Sept. 27, 2018, 4:20 p.m. UTC | #3
On Wed, Sep 26, 2018 at 09:56:07PM +0200, Goffredo Baroncelli wrote:
> On 25/09/2018 21.20, Daniel Kiper wrote:
> > On Wed, Sep 19, 2018 at 08:40:40PM +0200, Goffredo Baroncelli wrote:
> >> From: Goffredo Baroncelli <kreijack@inwind.it>
> >>
> [....]
> >>  	       *  - stripe_offset is the disk offset,
> >>  	       *  - csize is the "potential" data to read. It will be reduced to
> >>  	       *    size if the latter is smaller.
> >> +	       *  - parities_pos is the position of the parity inside a row (
> >
> > s/inside/in/>
> >> +	       *    2 for P1, 3 for P2...)
>
> +              *  - nparities is the number of parities (1 for RAID5, 2 for RAID6);
> +              *    used only in RAID5/6 code.
>
> >>  	       */
> >>  	      block_nr = grub_divmod64 (off, chunk_stripe_length, &low);
> >>
> >> @@ -1030,6 +1069,9 @@ grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
> >>  	       */
> >>  	      grub_divmod64 (high + stripen, nstripes, &stripen);
> >>
> >> +	      grub_divmod64 (high + nstripes - nparities, nstripes,
> >> +			     &parities_pos);
> >
> > I think that this math requires a bit of explanation in the comment
> > before grub_divmod64(). Especially I am interested in why high +
> > nstripes - nparities works as expected.
>
>
> What about
>
> /*
>  * parities_pos is equal to "(high - nparities) % nstripes" (see the diagram above).
>  * However "high - nparities" might be negative (eg when high == 0) leading to an
>  * incorrect computation.
>  * Instead "high + nstripes - nparities" is always positive and in modulo nstripes is
>  * equal to "(high - nparities) % nstripes
>  */

LGTM.

Daniel
diff mbox series

Patch

diff --git a/grub-core/fs/btrfs.c b/grub-core/fs/btrfs.c
index 55a7eeffc..400cd56b6 100644
--- a/grub-core/fs/btrfs.c
+++ b/grub-core/fs/btrfs.c
@@ -30,6 +30,7 @@ 
 #include <grub/i18n.h>
 #include <grub/btrfs.h>
 #include <grub/crypto.h>
+#include <grub/diskfilter.h>
 
 GRUB_MOD_LICENSE ("GPLv3+");
 
@@ -705,11 +706,36 @@  rebuild_raid5 (char *dest, struct raid56_buffer *buffers,
     }
 }
 
+static grub_err_t
+raid6_recover_read_buffer (void *data, int disk_nr,
+			   grub_uint64_t addr __attribute__ ((unused)),
+			   void *dest, grub_size_t size)
+{
+    struct raid56_buffer *buffers = data;
+
+    if (!buffers[disk_nr].data_is_valid)
+	return grub_errno = GRUB_ERR_READ_ERROR;
+
+    grub_memcpy(dest, buffers[disk_nr].buf, size);
+
+    return grub_errno = GRUB_ERR_NONE;
+}
+
+static void
+rebuild_raid6 (struct raid56_buffer *buffers, grub_uint64_t nstripes,
+               grub_uint64_t csize, grub_uint64_t parities_pos, void *dest,
+               grub_uint64_t stripen)
+
+{
+  grub_raid6_recover_gen (buffers, nstripes, stripen, parities_pos,
+                          dest, 0, csize, 0, raid6_recover_read_buffer);
+}
+
 static grub_err_t
 raid56_read_retry (struct grub_btrfs_data *data,
 		   struct grub_btrfs_chunk_item *chunk,
-		   grub_uint64_t stripe_offset,
-		   grub_uint64_t csize, void *buf)
+		   grub_uint64_t stripe_offset, grub_uint64_t stripen,
+		   grub_uint64_t csize, void *buf, grub_uint64_t parities_pos)
 {
   struct raid56_buffer *buffers;
   grub_uint64_t nstripes = grub_le_to_cpu16 (chunk->nstripes);
@@ -787,6 +813,15 @@  raid56_read_retry (struct grub_btrfs_data *data,
       ret = GRUB_ERR_READ_ERROR;
       goto cleanup;
     }
+  else if (failed_devices > 2 && (chunk_type & GRUB_BTRFS_CHUNK_TYPE_RAID6))
+    {
+      grub_dprintf ("btrfs",
+		    "not enough disks for raid6: total %" PRIuGRUB_UINT64_T
+		    ", missing %" PRIuGRUB_UINT64_T "\n",
+		    nstripes, failed_devices);
+      ret = GRUB_ERR_READ_ERROR;
+      goto cleanup;
+    }
   else
     grub_dprintf ("btrfs",
 		  "enough disks for RAID 5 rebuilding: total %"
@@ -797,7 +832,7 @@  raid56_read_retry (struct grub_btrfs_data *data,
   if (chunk_type & GRUB_BTRFS_CHUNK_TYPE_RAID5)
     rebuild_raid5 (buf, buffers, nstripes, csize);
   else
-    grub_dprintf ("btrfs", "called rebuild_raid6(), NOT IMPLEMENTED\n");
+    rebuild_raid6 (buffers, nstripes, csize, parities_pos, buf, stripen);
 
  cleanup:
   if (buffers)
@@ -886,9 +921,11 @@  grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
 	unsigned redundancy = 1;
 	unsigned i, j;
 	int is_raid56;
+	grub_uint64_t parities_pos = 0;
 
-	is_raid56 = !!(grub_le_to_cpu64 (chunk->type) &
-		       GRUB_BTRFS_CHUNK_TYPE_RAID5);
+        is_raid56 = !!(grub_le_to_cpu64 (chunk->type) &
+		       (GRUB_BTRFS_CHUNK_TYPE_RAID5 |
+		        GRUB_BTRFS_CHUNK_TYPE_RAID6));
 
 	if (grub_le_to_cpu64 (chunk->size) <= off)
 	  {
@@ -1015,6 +1052,8 @@  grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
 	       *  - stripe_offset is the disk offset,
 	       *  - csize is the "potential" data to read. It will be reduced to
 	       *    size if the latter is smaller.
+	       *  - parities_pos is the position of the parity inside a row (
+	       *    2 for P1, 3 for P2...)
 	       */
 	      block_nr = grub_divmod64 (off, chunk_stripe_length, &low);
 
@@ -1030,6 +1069,9 @@  grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
 	       */
 	      grub_divmod64 (high + stripen, nstripes, &stripen);
 
+	      grub_divmod64 (high + nstripes - nparities, nstripes,
+			     &parities_pos);
+
 	      stripe_offset = low + chunk_stripe_length * high;
 	      csize = chunk_stripe_length - low;
 
@@ -1081,7 +1123,7 @@  grub_btrfs_read_logical (struct grub_btrfs_data *data, grub_disk_addr_t addr,
 		grub_errno = GRUB_ERR_NONE;
 		if (err != GRUB_ERR_NONE)
 		  err = raid56_read_retry (data, chunk, stripe_offset,
-					   csize, buf);
+					   stripen, csize, buf, parities_pos);
 	      }
 	    if (err == GRUB_ERR_NONE)
 	      break;