From: "Stephen M. Cameron" <scameron@beardog.cce.hp.com>
To: linux-scsi@vger.kernel.org
Cc: James.Bottomley@HansenPartnership.com, dab@hp.com,
mikem@beardog.cce.hp.com
Subject: [PATCH 23/24] hpsa: fix block fetch table problem.
Date: Thu, 27 May 2010 15:14:39 -0500 [thread overview]
Message-ID: <20100527201439.3116.5339.stgit@beardog.cce.hp.com> (raw)
In-Reply-To: <20100527200301.3116.78973.stgit@beardog.cce.hp.com>
From: Stephen M. Cameron <scameron@beardog.cce.hp.com>
hpsa: fix block fetch table problem.
We have 32 (MAXSGENTRIES) scatter gather elements embedded
in the command. With all these, the total command size is
about 576 bytes. However, the last entry in the block fetch table
is 35. (the block fetch table contains the number of 16-byte chunks
the firmware needs to fetch for a given number of scatter gather
elements.) 35 * 16 = 560 bytes, which isn't enough. It needs to be
36. (36 * 16 == 576) or, MAXSGENTRIES + 4. (plus 4 because there's a
bunch of stuff at the front of the command before the first scatter
gather element that takes up 4 * 16 bytes.) Without this fix, the
controller may have to perform two DMA operations to fetch the
command since the first one may not get the whole thing.
Signed-off-by: Don Brace <brace@beardog.cce.hp.com>
Signed-off-by: Stephen M. Cameron <scameron@beardog.cce.hp.com>
---
drivers/scsi/hpsa.c | 21 ++++++++++++++++++++-
1 files changed, 20 insertions(+), 1 deletions(-)
diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c
index 066dde9..2c75f68 100644
--- a/drivers/scsi/hpsa.c
+++ b/drivers/scsi/hpsa.c
@@ -3834,7 +3834,26 @@ static __devinit void hpsa_enter_performant_mode(struct ctlr_info *h)
{
int i;
unsigned long register_value;
- int bft[8] = {5, 6, 8, 10, 12, 20, 28, 35}; /* for scatter/gathers */
+
+ /* This is a bit complicated. There are 8 registers on
+ * the controller which we write to to tell it 8 different
+ * sizes of commands which there may be. It's a way of
+ * reducing the DMA done to fetch each command. Encoded into
+ * each command's tag are 3 bits which communicate to the controller
+ * which of the eight sizes that command fits within. The size of
+ * each command depends on how many scatter gather entries there are.
+ * Each SG entry requires 16 bytes. The eight registers are programmed
+ * with the number of 16-byte blocks a command of that size requires.
+ * The smallest command possible requires 5 such 16 byte blocks.
+ * the largest command possible requires MAXSGENTRIES + 4 16-byte
+ * blocks. Note, this only extends to the SG entries contained
+ * within the command block, and does not extend to chained blocks
+ * of SG elements. bft[] contains the eight values we write to
+ * the registers. They are not evenly distributed, but have more
+ * sizes for small commands, and fewer sizes for larger commands.
+ */
+ int bft[8] = {5, 6, 8, 10, 12, 20, 28, MAXSGENTRIES + 4};
+ BUILD_BUG_ON(28 > MAXSGENTRIES + 4);
/* 5 = 1 s/g entry or 4k
* 6 = 2 s/g entry or 8k
* 8 = 4 s/g entry or 16k
next prev parent reply other threads:[~2010-05-27 20:10 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-27 20:12 [PATCH 00/24] hpsa updates for May 2010 Stephen M. Cameron
2010-05-27 20:12 ` [PATCH 01/24] hpsa: save pdev pointer in per hba structure early to avoid passing it around so much Stephen M. Cameron
2010-05-27 20:12 ` [PATCH 02/24] hpsa: factor out hpsa_lookup_board_id Stephen M. Cameron
2010-05-27 20:12 ` [PATCH 03/24] hpsa: factor out hpsa_board_disabled Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 04/24] hpsa: remove redundant board_id parameter from hpsa_interrupt_mode Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 05/24] hpsa: factor out hpsa_find_memory_BAR Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 06/24] hpsa: factor out hpsa_wait_for_board_ready Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 07/24] hpsa: factor out hpsa_find_cfgtables Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 08/24] hpsa: fix leak of ioremapped memory Stephen M. Cameron
2010-05-28 19:12 ` Rolf Eike Beer
2010-06-01 12:55 ` scameron
2010-05-27 20:13 ` [PATCH 09/24] hpsa: hpsa factor out hpsa_find_board_params Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 10/24] hpsa: factor out hpsa-CISS-signature-present Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 11/24] hpsa: factor out hpsa_enable_scsi_prefetch Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 12/24] hpsa: factor out hpsa_p600_dma_prefetch_quirk Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 13/24] hpsa: factor out hpsa_enter_simple_mode Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 14/24] hpsa: check that simple mode is supported Stephen M. Cameron
2010-05-27 20:13 ` [PATCH 15/24] hpsa: clean up debug ifdefs Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 16/24] hpsa: mark hpsa_mark_hpsa_put_ctlr_into_performant_mode as __devinit Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 17/24] hpsa: factor out hpsa_wait_for_mode_change_ack Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 18/24] hpsa: remove unused variable trans_offset Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 19/24] hpsa: factor out hpsa_enter_performant_mode Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 20/24] hpsa: remove unused firm_ver member of the per-hba structure Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 21/24] hpsa: Add hpsa.txt to Documentation/scsi Stephen M. Cameron
2010-05-27 20:14 ` [PATCH 22/24] hpsa: expose controller firmware revision via /sys Stephen M. Cameron
2010-05-27 20:14 ` Stephen M. Cameron [this message]
2010-05-27 20:14 ` [PATCH 24/24] hpsa: add entry to MAINTAINERS Stephen M. Cameron
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=20100527201439.3116.5339.stgit@beardog.cce.hp.com \
--to=scameron@beardog.cce.hp.com \
--cc=James.Bottomley@HansenPartnership.com \
--cc=dab@hp.com \
--cc=linux-scsi@vger.kernel.org \
--cc=mikem@beardog.cce.hp.com \
/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).