From: Mark Lord <liml@rtr.ca>
To: Jeff Garzik <jgarzik@pobox.com>
Cc: linux-ide@vger.kernel.org
Subject: [PATCH 5/7] sata_mv: remove local copy of queue indexes
Date: Fri, 19 May 2006 16:36:36 -0400 [thread overview]
Message-ID: <200605191636.36648.liml@rtr.ca> (raw)
In-Reply-To: <200605191614.00023.liml@rtr.ca>
The driver currently keeps local copies of the hardware request/response queue indexes.
But it expends significant effort ensuring consistency between the two views,
and still gets it wrong after an error or reset occurs.
This patch removes the local copies, in favour of just accessing the hardware
whenever we need them. Eventually this may need to be tweaked again for NCQ,
but for now this works and solves problems some users were seeing.
Signed-off-by: Mark Lord <liml@rtr.ca>
---
--- linux/drivers/scsi/sata_mv.c 2006-05-19 15:40:46.000000000 -0400
+++ linux/drivers/scsi/sata_mv.c 2006-05-19 15:44:11.000000000 -0400
@@ -308,9 +308,6 @@
dma_addr_t crpb_dma;
struct mv_sg *sg_tbl;
dma_addr_t sg_tbl_dma;
-
- unsigned req_producer; /* cp of req_in_ptr */
- unsigned rsp_consumer; /* cp of rsp_out_ptr */
u32 pp_flags;
};
@@ -943,8 +940,6 @@
writelfl(pp->crpb_dma & EDMA_RSP_Q_BASE_LO_MASK,
port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
- pp->req_producer = pp->rsp_consumer = 0;
-
/* Don't turn on EDMA here...do it before DMA commands only. Else
* we'll be unable to send non-data, PIO, etc due to restricted access
* to shadow regs.
@@ -1028,10 +1023,9 @@
}
}
-static inline unsigned mv_inc_q_index(unsigned *index)
+static inline unsigned mv_inc_q_index(unsigned index)
{
- *index = (*index + 1) & MV_MAX_Q_DEPTH_MASK;
- return *index;
+ return (index + 1) & MV_MAX_Q_DEPTH_MASK;
}
static inline void mv_crqb_pack_cmd(u16 *cmdw, u8 data, u8 addr, unsigned last)
@@ -1059,15 +1053,11 @@
u16 *cw;
struct ata_taskfile *tf;
u16 flags = 0;
+ unsigned in_index;
if (ATA_PROT_DMA != qc->tf.protocol)
return;
- /* the req producer index should be the same as we remember it */
- WARN_ON(((readl(mv_ap_base(qc->ap) + EDMA_REQ_Q_IN_PTR_OFS) >>
- EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) !=
- pp->req_producer);
-
/* Fill in command request block
*/
if (!(qc->tf.flags & ATA_TFLAG_WRITE))
@@ -1075,13 +1065,17 @@
WARN_ON(MV_MAX_Q_DEPTH <= qc->tag);
flags |= qc->tag << CRQB_TAG_SHIFT;
- pp->crqb[pp->req_producer].sg_addr =
+ /* get current queue index from hardware */
+ in_index = (readl(mv_ap_base(ap) + EDMA_REQ_Q_IN_PTR_OFS)
+ >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK;
+
+ pp->crqb[in_index].sg_addr =
cpu_to_le32(pp->sg_tbl_dma & 0xffffffff);
- pp->crqb[pp->req_producer].sg_addr_hi =
+ pp->crqb[in_index].sg_addr_hi =
cpu_to_le32((pp->sg_tbl_dma >> 16) >> 16);
- pp->crqb[pp->req_producer].ctrl_flags = cpu_to_le16(flags);
+ pp->crqb[in_index].ctrl_flags = cpu_to_le16(flags);
- cw = &pp->crqb[pp->req_producer].ata_cmd[0];
+ cw = &pp->crqb[in_index].ata_cmd[0];
tf = &qc->tf;
/* Sadly, the CRQB cannot accomodate all registers--there are
@@ -1150,16 +1144,12 @@
struct mv_port_priv *pp = ap->private_data;
struct mv_crqb_iie *crqb;
struct ata_taskfile *tf;
+ unsigned in_index;
u32 flags = 0;
if (ATA_PROT_DMA != qc->tf.protocol)
return;
- /* the req producer index should be the same as we remember it */
- WARN_ON(((readl(mv_ap_base(qc->ap) + EDMA_REQ_Q_IN_PTR_OFS) >>
- EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) !=
- pp->req_producer);
-
/* Fill in Gen IIE command request block
*/
if (!(qc->tf.flags & ATA_TFLAG_WRITE))
@@ -1168,7 +1158,11 @@
WARN_ON(MV_MAX_Q_DEPTH <= qc->tag);
flags |= qc->tag << CRQB_TAG_SHIFT;
- crqb = (struct mv_crqb_iie *) &pp->crqb[pp->req_producer];
+ /* get current queue index from hardware */
+ in_index = (readl(mv_ap_base(ap) + EDMA_REQ_Q_IN_PTR_OFS)
+ >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK;
+
+ crqb = (struct mv_crqb_iie *) &pp->crqb[in_index];
crqb->addr = cpu_to_le32(pp->sg_tbl_dma & 0xffffffff);
crqb->addr_hi = cpu_to_le32((pp->sg_tbl_dma >> 16) >> 16);
crqb->flags = cpu_to_le32(flags);
@@ -1216,6 +1210,7 @@
{
void __iomem *port_mmio = mv_ap_base(qc->ap);
struct mv_port_priv *pp = qc->ap->private_data;
+ unsigned in_index;
u32 in_ptr;
if (ATA_PROT_DMA != qc->tf.protocol) {
@@ -1227,23 +1222,20 @@
return ata_qc_issue_prot(qc);
}
- in_ptr = readl(port_mmio + EDMA_REQ_Q_IN_PTR_OFS);
+ in_ptr = readl(port_mmio + EDMA_REQ_Q_IN_PTR_OFS);
+ in_index = (in_ptr >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK;
- /* the req producer index should be the same as we remember it */
- WARN_ON(((in_ptr >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) !=
- pp->req_producer);
/* until we do queuing, the queue should be empty at this point */
- WARN_ON(((in_ptr >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) !=
- ((readl(port_mmio + EDMA_REQ_Q_OUT_PTR_OFS) >>
- EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK));
+ WARN_ON(in_index != ((readl(port_mmio + EDMA_REQ_Q_OUT_PTR_OFS)
+ >> EDMA_REQ_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK));
- mv_inc_q_index(&pp->req_producer); /* now incr producer index */
+ in_index = mv_inc_q_index(in_index); /* now incr producer index */
mv_start_dma(port_mmio, pp);
/* and write the request in pointer to kick the EDMA to life */
in_ptr &= EDMA_REQ_Q_BASE_LO_MASK;
- in_ptr |= pp->req_producer << EDMA_REQ_Q_PTR_SHIFT;
+ in_ptr |= in_index << EDMA_REQ_Q_PTR_SHIFT;
writelfl(in_ptr, port_mmio + EDMA_REQ_Q_IN_PTR_OFS);
return 0;
@@ -1266,28 +1258,26 @@
{
void __iomem *port_mmio = mv_ap_base(ap);
struct mv_port_priv *pp = ap->private_data;
+ unsigned out_index;
u32 out_ptr;
u8 ata_status;
- out_ptr = readl(port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
-
- /* the response consumer index should be the same as we remember it */
- WARN_ON(((out_ptr >> EDMA_RSP_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) !=
- pp->rsp_consumer);
+ out_ptr = readl(port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
+ out_index = (out_ptr >> EDMA_RSP_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK;
- ata_status = pp->crpb[pp->rsp_consumer].flags >> CRPB_FLAG_STATUS_SHIFT;
+ ata_status = le16_to_cpu(pp->crpb[out_index].flags)
+ >> CRPB_FLAG_STATUS_SHIFT;
/* increment our consumer index... */
- pp->rsp_consumer = mv_inc_q_index(&pp->rsp_consumer);
+ out_index = mv_inc_q_index(out_index);
/* and, until we do NCQ, there should only be 1 CRPB waiting */
- WARN_ON(((readl(port_mmio + EDMA_RSP_Q_IN_PTR_OFS) >>
- EDMA_RSP_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK) !=
- pp->rsp_consumer);
+ WARN_ON(out_index != ((readl(port_mmio + EDMA_RSP_Q_IN_PTR_OFS)
+ >> EDMA_RSP_Q_PTR_SHIFT) & MV_MAX_Q_DEPTH_MASK));
/* write out our inc'd consumer index so EDMA knows we're caught up */
out_ptr &= EDMA_RSP_Q_BASE_LO_MASK;
- out_ptr |= pp->rsp_consumer << EDMA_RSP_Q_PTR_SHIFT;
+ out_ptr |= out_index << EDMA_RSP_Q_PTR_SHIFT;
writelfl(out_ptr, port_mmio + EDMA_RSP_Q_OUT_PTR_OFS);
/* Return ATA status register for completed CRPB */
next prev parent reply other threads:[~2006-05-19 20:36 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-19 15:48 [PATCHSET 03/03] add hotplug support, take 3 Tejun Heo
2006-05-19 15:48 ` [PATCH 04/13] libata-hp: implement warmplug Tejun Heo
2006-05-19 15:48 ` [PATCH 08/13] ata_piix: convert ata_piix to new probing mechanism Tejun Heo
2006-05-19 15:48 ` [PATCH 03/13] libata-hp: implement SCSI part of hotplug Tejun Heo
2006-05-19 16:05 ` Jeff Garzik
2006-05-23 14:52 ` Tejun Heo
2006-05-19 15:48 ` [PATCH 01/13] libata-hp: implement ata_eh_detach_dev() Tejun Heo
2006-05-19 15:48 ` [PATCH 06/13] libata-hp: implement bootplug Tejun Heo
2006-05-19 16:09 ` Jeff Garzik
2006-05-19 15:48 ` [PATCH 09/13] sata_sil: convert to new probing mechanism and add hotplug support Tejun Heo
2006-05-19 15:48 ` [PATCH 05/13] libata-hp: hook warmplug Tejun Heo
2006-05-19 15:48 ` [PATCH 02/13] libata-hp: implement hotplug Tejun Heo
2006-05-19 16:04 ` Jeff Garzik
2006-05-19 15:48 ` [PATCH 07/13] libata-hp: implement unload-unplug Tejun Heo
2006-05-19 16:10 ` Jeff Garzik
2006-05-23 14:53 ` Tejun Heo
2006-05-19 15:48 ` [PATCH 12/13] libata-hp: killl ops->probe_reset Tejun Heo
2006-05-19 15:48 ` [PATCH 10/13] ahci: convert to new probing mechanism and add hotplug support Tejun Heo
2006-05-19 15:48 ` [PATCH 11/13] sata_sil24: " Tejun Heo
2006-05-19 15:48 ` [PATCH 13/13] libata-hp: move ata_do_reset() to libata-eh.c Tejun Heo
2006-05-19 16:13 ` Jeff Garzik
2006-05-19 20:13 ` [PATCH 0/7] sata_mv: assorted fixes Mark Lord
2006-05-19 20:21 ` [PATCH 1/7] sata_mv: prevent unnecessary double-resets Mark Lord
2006-05-19 20:24 ` [PATCH 2/7] sata_mv: deal with interrupt coalescing interrupts Mark Lord
2006-05-20 4:32 ` Jeff Garzik
2006-05-20 13:13 ` Mark Lord
2006-05-20 13:24 ` Jeff Garzik
2006-05-20 14:01 ` Mark Lord
2006-05-22 6:35 ` Jeff Garzik
2006-05-19 20:29 ` [PATCH 3/7] sata_mv: chip initialization fixes Mark Lord
2006-05-19 20:33 ` [PATCH 4/7] sata_mv: spurious interrupt workaround Mark Lord
2006-05-19 20:36 ` Mark Lord [this message]
2006-05-19 20:40 ` [PATCH 6/7] sata_mv: endian fix Mark Lord
2006-05-19 20:41 ` [PATCH 7/7] sata_mv: version bump Mark Lord
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=200605191636.36648.liml@rtr.ca \
--to=liml@rtr.ca \
--cc=jgarzik@pobox.com \
--cc=linux-ide@vger.kernel.org \
/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).