Skip to contentSkip to conversation

libata: make parallel port scan a per-host opt-in feature

PATCH New Arjan van de Ven 2 participants 7 messages Jan 10, 2009 libata

lore
Share & tools
lore
#0 lore raw
Quoted 1 line

From 17f458c7d9d7a25a04d448aef22b134135857018 Mon Sep 17 00:00:00 2001

From: Arjan van de Ven <[email protected]>
Date: Fri, 9 Jan 2009 15:54:07 -0800
Subject: [PATCH] libata: Add a per-host flag to opt-in into parallel port probes

This patch adds a per host flag that allows drivers to opt in into having its busses scanned in parallel.

Drivers that do not set this flag get their ports scanned in the "original" sequence.

---
 drivers/ata/ahci.c        |    3 +++
 drivers/ata/libata-core.c |    9 +++++++++
 include/linux/libata.h    |    1 +
 3 files changed, 13 insertions(+), 0 deletions(-)
drivers/ata/ahci.c +3 -0
index 7f701cb..1fc2d4b 100644
@@ -2660,6 +2660,9 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	host->iomap = pcim_iomap_table(pdev);
 	host->private_data = hpriv;
 
+	if (!hpriv->cap & HOST_CAP_SSS)
+		host->flags |= ATA_HOST_PARALLEL_SCAN;
+
 	if (pi.flags & ATA_FLAG_EM)
 		ahci_reset_em(host);
 
drivers/ata/libata-core.c +9 -0
index c507a9a..31c0f22 100644
@@ -5920,6 +5920,15 @@ static void async_port_probe(void *data, async_cookie_t cookie)
 {
 	int rc;
 	struct ata_port *ap = data;
+
+	/*
+	 * If we're not allowed to scan this host in parallel,
+	 * we need to wait until all previous scans have completed
+	 * before going further.
+	 */
+	if (!(ap->host->flags & ATA_HOST_PARALLEL_SCAN))
+		async_synchronize_cookie(cookie);
+	
 	/* probe */
 	if (ap->ops->error_handler) {
 		struct ata_eh_info *ehi = &ap->link.eh_info;
include/linux/libata.h +1 -0
index 4f7c8fb..b6b8a7f 100644
@@ -239,6 +239,7 @@ enum {
 	/* host set flags */
 	ATA_HOST_SIMPLEX	= (1 << 0),	/* Host is simplex, one DMA channel per host only */
 	ATA_HOST_STARTED	= (1 << 1),	/* Host started */
+	ATA_HOST_PARALLEL_SCAN	= (1 << 2),	/* Ports on this host can be scanned in parallel */
 
 	/* bits 24:31 of host->flags are reserved for LLD specific flags */
 
Signature · 7 lines
1.6.0.6


-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org
#1 lore raw

drivers/ata/ahci.c static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) · line 2665

Quoted 4 lines · Arjan van de Ven
drivers/ata/ahci.c +3 -0
+	if (!hpriv->cap & HOST_CAP_SSS)
+		host->flags |= ATA_HOST_PARALLEL_SCAN;
+

I don't think that does what you think it does.

Linus

#2 lore raw
Quoted 9 lines · Linus Torvalds

drivers/ata/ahci.c static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) · line 2665

Quoted 4 lines · Arjan van de Ven
drivers/ata/ahci.c +3 -0
+	if (!hpriv->cap & HOST_CAP_SSS)
+		host->flags |= ATA_HOST_PARALLEL_SCAN;
+

I don't think that does what you think it does.

And furthermore I'm surprised gcc doesn't warn about it. It would seem to be an obvious thing to warn about. I'll see what a sparse warning would look like.

Linus

#3 lore raw
Quoted 11 lines · Linus Torvalds
Quoted 8 lines · Linus Torvalds

drivers/ata/ahci.c static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) · line 2665

Quoted 4 lines · Arjan van de Ven
drivers/ata/ahci.c +3 -0
+	if (!hpriv->cap & HOST_CAP_SSS)
+		host->flags |= ATA_HOST_PARALLEL_SCAN;
+

I don't think that does what you think it does.

Quoted 3 lines

And furthermore I'm surprised gcc doesn't warn about it. It would seem to be an obvious thing to warn about. I'll see what a sparse warning would look like.

Heh. sparse already warns:

warning: dubious: !x & y

although I'm not sure that's necessarily the most readable warning ever.

Linus

#4 lore raw
In reply to Linus Torvalds
Quoted 11 lines · On Fri, 9 Jan 2009 16:56:59 -0800 (PST) Linus Torvalds…

drivers/ata/ahci.c static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) · line 2665

Quoted 4 lines · Arjan van de Ven
drivers/ata/ahci.c +3 -0
+	if (!hpriv->cap & HOST_CAP_SSS)
+		host->flags |= ATA_HOST_PARALLEL_SCAN;
+

I don't think that does what you think it does.

Linus

argh two few ()'s

Signature · 3 lines
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org
#5 lore raw

PATCH 1/2 libata: Add a per-host flag to opt-in into parallel port probes

In reply to Linus Torvalds
Quoted 1 line

From 703d035d397fb2a253559f3b819ad44702807724 Mon Sep 17 00:00:00 2001

From: Arjan van de Ven <[email protected]>
Date: Fri, 9 Jan 2009 15:54:07 -0800
Subject: [PATCH] libata: Add a per-host flag to opt-in into parallel port probes

This patch adds a per host flag that allows drivers to opt in into having its busses scanned in parallel.

Drivers that do not set this flag get their ports scanned in the "original" sequence.

---
 drivers/ata/ahci.c        |    3 +++
 drivers/ata/libata-core.c |    9 +++++++++
 include/linux/libata.h    |    1 +
 3 files changed, 13 insertions(+), 0 deletions(-)
drivers/ata/ahci.c +3 -0
index 7f701cb..9603967 100644
@@ -2660,6 +2660,9 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	host->iomap = pcim_iomap_table(pdev);
 	host->private_data = hpriv;
 
+	if (!(hpriv->cap & HOST_CAP_SSS))
+		host->flags |= ATA_HOST_PARALLEL_SCAN;
+
 	if (pi.flags & ATA_FLAG_EM)
 		ahci_reset_em(host);
 
drivers/ata/libata-core.c +9 -0
index c507a9a..31c0f22 100644
@@ -5920,6 +5920,15 @@ static void async_port_probe(void *data, async_cookie_t cookie)
 {
 	int rc;
 	struct ata_port *ap = data;
+
+	/*
+	 * If we're not allowed to scan this host in parallel,
+	 * we need to wait until all previous scans have completed
+	 * before going further.
+	 */
+	if (!(ap->host->flags & ATA_HOST_PARALLEL_SCAN))
+		async_synchronize_cookie(cookie);
+	
 	/* probe */
 	if (ap->ops->error_handler) {
 		struct ata_eh_info *ehi = &ap->link.eh_info;
include/linux/libata.h +1 -0
index 4f7c8fb..b6b8a7f 100644
@@ -239,6 +239,7 @@ enum {
 	/* host set flags */
 	ATA_HOST_SIMPLEX	= (1 << 0),	/* Host is simplex, one DMA channel per host only */
 	ATA_HOST_STARTED	= (1 << 1),	/* Host started */
+	ATA_HOST_PARALLEL_SCAN	= (1 << 2),	/* Ports on this host can be scanned in parallel */
 
 	/* bits 24:31 of host->flags are reserved for LLD specific flags */
 
Signature · 8 lines
1.6.0.6



-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org
#6 lore raw

PATCH 2/2 libata: only ports >= 0 need to synchronize

Quoted 1 line

From 68db7c1eff38c2e121ccc922fd365f4f086369dc Mon Sep 17 00:00:00 2001

From: Arjan van de Ven <[email protected]>
Date: Sat, 10 Jan 2009 10:18:44 -0800
Subject: [PATCH] libata: only ports >= 0 need to synchronize

In a discussio with Jeff Garzik, he mentioned that the serialization for the libata port probes only needs to be within the domain of a host. This means that for the first port of each host (with ID 0), we don't need to wait, so we can relax our serialization a little.

---
 drivers/ata/libata-core.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)
drivers/ata/libata-core.c +4 -2
index 31c0f22..71218d7 100644
@@ -5925,10 +5925,12 @@ static void async_port_probe(void *data, async_cookie_t cookie)
 	 * If we're not allowed to scan this host in parallel,
 	 * we need to wait until all previous scans have completed
 	 * before going further.
+	 * Jeff Garzik says this is only within a controller, so we
+	 * don't need to wait for port 0, only for later ports.
 	 */
-	if (!(ap->host->flags & ATA_HOST_PARALLEL_SCAN))
+	if (!(ap->host->flags & ATA_HOST_PARALLEL_SCAN) && ap->port_no != 0)
 		async_synchronize_cookie(cookie);
-	
+
 	/* probe */
 	if (ap->ops->error_handler) {
 		struct ata_eh_info *ehi = &ap->link.eh_info;
Signature · 8 lines
1.6.0.6



-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org