Skip to contentSkip to conversation

mvsata_ide: adjust port init sequence

PATCH New Albert Aribaud 2 participants 4 messages Sep 6, 2010 – Sep 7, 2010 Block layer

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mvsata_ide: adjust port init sequence

mvsata_ide_initialize_port(): adjust init sequence (SStatus should be checked only after all writes to SControl) and return success/failure to ide_preinit().

Also, as some tests showed init durations in the hundreds of us, raise the time-out to 10 ms to be on the safe side.

---
PATCH HISTORY

V1 Initial patch.
V2 Fixed wrong placement of comment.
Fixed wrong description (was 01 ms, should have been 10). Added Signed-off-by.

 drivers/block/mvsata_ide.c |   22 ++++++++++++++--------
 1 files changed, 14 insertions(+), 8 deletions(-)
drivers/block/mvsata_ide.c +14 -8
index 077b278..470659d 100644
@@ -97,23 +97,27 @@ struct mvsata_port_registers {
  * DET back to "no action".
  */
 
-static void mvsata_ide_initialize_port(struct mvsata_port_registers *port)
+static int mvsata_ide_initialize_port(struct mvsata_port_registers *port)
 {
 	u32 control;
 	u32 status;
-	u32 tout = 50; /* wait at most 50 us for SATA reset to complete */
+	u32 tout = 10000; /* wait at most 10 ms for SATA reset to complete */
 
+	/* Set control IPM to 3 (no low power) and DET to 1 (initialize) */
 	control = readl(&port->scontrol);
 	control = (control & ~MVSATA_SCONTROL_MASK) | MVSATA_PORT_INIT;
 	writel(control, &port->scontrol);
+	/* Toggle control DET back to 0 (normal operation) */
+	control = (control & ~MVSATA_SCONTROL_MASK) | MVSATA_PORT_USE;
+	writel(control, &port->scontrol);
+	/* wait for status DET to become 3 (device and communication OK) */
 	while (--tout) {
 		status = readl(&port->sstatus) & MVSATA_SSTATUS_DET_MASK;
 		if (status == MVSATA_SSTATUS_DET_DEVCOMM)
 			break;
 		udelay(1);
 	}
-	control = (control & ~MVSATA_SCONTROL_MASK) | MVSATA_PORT_USE;
-	writel(control, &port->scontrol);
+	return (tout? 0: 1);
 }
 
 /*
@@ -125,15 +129,17 @@ int ide_preinit(void)
 {
 	/* Enable ATA port 0 (could be SATA port 0 or 1) if declared */
 #if defined(CONFIG_SYS_ATA_IDE0_OFFSET)
-	mvsata_ide_initialize_port(
+	if (mvsata_ide_initialize_port(
 		(struct mvsata_port_registers *)
-		(CONFIG_SYS_ATA_BASE_ADDR + CONFIG_SYS_ATA_IDE0_OFFSET));
+		(CONFIG_SYS_ATA_BASE_ADDR + CONFIG_SYS_ATA_IDE0_OFFSET)))
+		return 1;
 #endif
 	/* Enable ATA port 1 (could be SATA port 0 or 1) if declared */
 #if defined(CONFIG_SYS_ATA_IDE1_OFFSET)
-	mvsata_ide_initialize_port(
+	if (mvsata_ide_initialize_port(
 		(struct mvsata_port_registers *)
-		(CONFIG_SYS_ATA_BASE_ADDR + CONFIG_SYS_ATA_IDE1_OFFSET));
+		(CONFIG_SYS_ATA_BASE_ADDR + CONFIG_SYS_ATA_IDE1_OFFSET)))
+		return 1;
 #endif
 	/* return 0 as we always succeed */
 	return 0;
Signature · 1 line
1.7.0.4
#1 lore raw

PATCH 1/1 mvsata_ide: adjust port init sequence

Le 07/09/2010 00:53, Albert Aribaud a ?crit :

Quoted 1 line

Sorry, missed the "V2" part in the tag -- ignore this post.

Amicalement,

Signature · 1 line
Albert.
#2 lore raw

PATCH 1/1 mvsata_ide: adjust port init sequence

Dear Albert ARIBAUD,

Quoted 4 lines · In message <[email protected]> you wrote:

Le 07/09/2010 00:53, Albert Aribaud a =E9crit :

Quoted 1 line

Sorry, missed the "V2" part in the tag -- ignore this post.

Argh... And I missed the followup.

Please ignore my mail, too. :-(

Best regards,

Wolfgang Denk

Signature · 5 lines
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
"If that makes any sense to you, you have a big problem."
                                  -- C. Durance, Computer Science 234
#3 lore raw

PATCH 1/1 mvsata_ide: adjust port init sequence

In reply to Albert Aribaud

Dear Albert Aribaud,

Quoted 7 lines · In message <[email protected]> you wrote:

mvsata_ide_initialize_port(): adjust init sequence (SStatus should be checked only after all writes to SControl) and return success/failure to ide_preinit().

Also, as some tests showed init durations in the hundreds of us, raise the time-out to 10 ms to be on the safe side.

Quoted 8 lines

---
PATCH HISTORY

V1 Initial patch.
V2 Fixed wrong placement of comment.
Fixed wrong description (was 01 ms, should have been 10). Added Signed-off-by.

Could you _please_ include the patch version information in the
Subject: line?

Thanks.

Note: the "--subject-prefix" option for git-format-patch allows to do this easily, without manual editing of the patch(es).

Best regards,

Wolfgang Denk

Signature · 5 lines
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
A good marriage would be between a blind wife and deaf husband.
                                               -- Michel de Montaigne