Yuri Tikhonov wrote:
This patch implements support for the asynchronous computation of RAID-6
syndromes.
It provides an API to compute RAID-6 syndromes asynchronously in a format
conforming to async_tx interfaces. The async_pxor and async_pqxor_zero_sum
functions are very similar to async_xor functions but make use of
additional tx_set_src_mult method for setting cooefficients of the RAID-6
Q syndrome.
The Galois polynomial which is used in the s/w case is 0x11d (the
corresponding coefficients are hard-coded in raid6_call.gen_syndrome).
Because even with the h/w acceleration enabled some pqxor operations may be
processed in CPU (e.g. in case of no DMA descriptors available) it's highly
recommended to configure the DMA engine which your system uses to exploit
exactly the same Galois polynomial.
It should probably be noted here, too, that if you use a different basis
polynomial for the Galois field you will end up with a different on-disk
format.
+ * You should have received a copy of the GNU General Public License
along with
+ * this program; if not, write to the Free Software Foundation, Inc., 59
+ * Temple Place - Suite 330, Boston, MA 02111-1307, USA.
This address, I believe, is obsolete.
+ if (!(tx=async_pqxor(NULL, ptrs[failb],
+ &ptrs[disks - 2], bc, 0, 2, bytes,
+ ASYNC_TX_DEP_ACK | ASYNC_TX_XOR_ZERO_DST,
+ tx, NULL, NULL))) {
+ /* It's bad if we failed here; try to repeat this
+ * using another failed disk as a spare; this wouldn't
+ * failed since now we'll be able to compute synchronously
+ * (there is no support for synchronous Q-only)
+ */
+ async_pqxor(ptrs[faila], ptrs[failb],
+ &ptrs[disks - 2], bc, 0, 2, bytes,
+ ASYNC_TX_DEP_ACK | ASYNC_TX_XOR_ZERO_DST,
+ NULL, NULL, NULL);
+ }
I don't really understand this logic, or the comment that goes along
with it. Could you please elucidate?
-hpa
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
