diff options
author | Dan Williams <dan.j.williams@intel.com> | 2009-09-08 20:55:21 -0400 |
---|---|---|
committer | Dan Williams <dan.j.williams@intel.com> | 2009-09-08 20:55:21 -0400 |
commit | bbb20089a3275a19e475dbc21320c3742e3ca423 (patch) | |
tree | 216fdc1cbef450ca688135c5b8969169482d9a48 /Documentation/mtd/nand_ecc.txt | |
parent | 3e48e656903e9fd8bc805c6a2c4264d7808d315b (diff) | |
parent | 657a77fa7284d8ae28dfa48f1dc5d919bf5b2843 (diff) |
Merge branch 'dmaengine' into async-tx-next
Conflicts:
crypto/async_tx/async_xor.c
drivers/dma/ioat/dma_v2.h
drivers/dma/ioat/pci.c
drivers/md/raid5.c
Diffstat (limited to 'Documentation/mtd/nand_ecc.txt')
-rw-r--r-- | Documentation/mtd/nand_ecc.txt | 12 |
1 files changed, 6 insertions, 6 deletions
diff --git a/Documentation/mtd/nand_ecc.txt b/Documentation/mtd/nand_ecc.txt index bdf93b7f0f24..274821b35a7f 100644 --- a/Documentation/mtd/nand_ecc.txt +++ b/Documentation/mtd/nand_ecc.txt | |||
@@ -50,7 +50,7 @@ byte 255: bit7 bit6 bit5 bit4 bit3 bit2 bit1 bit0 rp1 rp3 rp5 ... rp15 | |||
50 | cp5 cp5 cp5 cp5 cp4 cp4 cp4 cp4 | 50 | cp5 cp5 cp5 cp5 cp4 cp4 cp4 cp4 |
51 | 51 | ||
52 | This figure represents a sector of 256 bytes. | 52 | This figure represents a sector of 256 bytes. |
53 | cp is my abbreviaton for column parity, rp for row parity. | 53 | cp is my abbreviation for column parity, rp for row parity. |
54 | 54 | ||
55 | Let's start to explain column parity. | 55 | Let's start to explain column parity. |
56 | cp0 is the parity that belongs to all bit0, bit2, bit4, bit6. | 56 | cp0 is the parity that belongs to all bit0, bit2, bit4, bit6. |
@@ -560,7 +560,7 @@ Measuring this code again showed big gain. When executing the original | |||
560 | linux code 1 million times, this took about 1 second on my system. | 560 | linux code 1 million times, this took about 1 second on my system. |
561 | (using time to measure the performance). After this iteration I was back | 561 | (using time to measure the performance). After this iteration I was back |
562 | to 0.075 sec. Actually I had to decide to start measuring over 10 | 562 | to 0.075 sec. Actually I had to decide to start measuring over 10 |
563 | million interations in order not to loose too much accuracy. This one | 563 | million iterations in order not to lose too much accuracy. This one |
564 | definitely seemed to be the jackpot! | 564 | definitely seemed to be the jackpot! |
565 | 565 | ||
566 | There is a little bit more room for improvement though. There are three | 566 | There is a little bit more room for improvement though. There are three |
@@ -571,8 +571,8 @@ loop; This eliminates 3 statements per loop. Of course after the loop we | |||
571 | need to correct by adding: | 571 | need to correct by adding: |
572 | rp4 ^= rp4_6; | 572 | rp4 ^= rp4_6; |
573 | rp6 ^= rp4_6 | 573 | rp6 ^= rp4_6 |
574 | Furthermore there are 4 sequential assingments to rp8. This can be | 574 | Furthermore there are 4 sequential assignments to rp8. This can be |
575 | encoded slightly more efficient by saving tmppar before those 4 lines | 575 | encoded slightly more efficiently by saving tmppar before those 4 lines |
576 | and later do rp8 = rp8 ^ tmppar ^ notrp8; | 576 | and later do rp8 = rp8 ^ tmppar ^ notrp8; |
577 | (where notrp8 is the value of rp8 before those 4 lines). | 577 | (where notrp8 is the value of rp8 before those 4 lines). |
578 | Again a use of the commutative property of xor. | 578 | Again a use of the commutative property of xor. |
@@ -622,7 +622,7 @@ Not a big change, but every penny counts :-) | |||
622 | Analysis 7 | 622 | Analysis 7 |
623 | ========== | 623 | ========== |
624 | 624 | ||
625 | Acutally this made things worse. Not very much, but I don't want to move | 625 | Actually this made things worse. Not very much, but I don't want to move |
626 | into the wrong direction. Maybe something to investigate later. Could | 626 | into the wrong direction. Maybe something to investigate later. Could |
627 | have to do with caching again. | 627 | have to do with caching again. |
628 | 628 | ||
@@ -642,7 +642,7 @@ Analysis 8 | |||
642 | This makes things worse. Let's stick with attempt 6 and continue from there. | 642 | This makes things worse. Let's stick with attempt 6 and continue from there. |
643 | Although it seems that the code within the loop cannot be optimised | 643 | Although it seems that the code within the loop cannot be optimised |
644 | further there is still room to optimize the generation of the ecc codes. | 644 | further there is still room to optimize the generation of the ecc codes. |
645 | We can simply calcualate the total parity. If this is 0 then rp4 = rp5 | 645 | We can simply calculate the total parity. If this is 0 then rp4 = rp5 |
646 | etc. If the parity is 1, then rp4 = !rp5; | 646 | etc. If the parity is 1, then rp4 = !rp5; |
647 | But if rp4 = rp5 we do not need rp5 etc. We can just write the even bits | 647 | But if rp4 = rp5 we do not need rp5 etc. We can just write the even bits |
648 | in the result byte and then do something like | 648 | in the result byte and then do something like |