2011-04-23 06:58:29

by Jian Peng

Subject: [PATCH 1/1] libata: ahci_start_engine compliant to AHCI spec

At the end of section 10.1 of AHCI spec (rev 1.3), it states

Software shall not set PxCMD.ST to 1 until it is determined that
a functoinal device is present on the port as determined by
PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h

Even though most AHCI host controller works without this check,
specific controller will fail under this condition.

Signed-off-by: Jian Peng <[email protected]>
---
drivers/ata/libahci.c | 21 +++++++++++++++++++++
1 files changed, 21 insertions(+), 0 deletions(-)

diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index 26d4523..83ed544 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -539,6 +539,27 @@ void ahci_start_engine(struct ata_port *ap)
{
void __iomem *port_mmio = ahci_port_base(ap);
u32 tmp;
+ u8 status;
+
+ status = readl(port_mmio + PORT_TFDATA) & 0xFF;
+
+ /*
+ * At end of section 10.1 of AHCI spec (rev 1.3), it states
+ * Software shall not set PxCMD.ST to 1 until it is determined
+ * that a functoinal device is present on the port as determined by
+ * PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h
+ *
+ * Even though most AHCI host controllers work without this check,
+ * specific controller will fail under this condition
+ */
+ if (status & (ATA_BUSY | ATA_DRQ))
+ return;
+ else {
+ ahci_scr_read(&ap->link, SCR_STATUS, &tmp);
+
+ if ((tmp & 0xf) != 0x3)
+ return;
+ }

/* start DMA */
tmp = readl(port_mmio + PORT_CMD);
--
1.7.4.1

2011-04-23 07:08:12

by Jeff Garzik

Subject: Re: [PATCH 1/1] libata: ahci_start_engine compliant to AHCI spec

On 04/23/2011 02:58 AM, Jian Peng wrote:
> At the end of section 10.1 of AHCI spec (rev 1.3), it states
>
> Software shall not set PxCMD.ST to 1 until it is determined that
> a functoinal device is present on the port as determined by
> PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h
>
> Even though most AHCI host controller works without this check,
> specific controller will fail under this condition.
>
> Signed-off-by: Jian Peng<[email protected]>
> ---
> drivers/ata/libahci.c | 21 +++++++++++++++++++++
> 1 files changed, 21 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
> index 26d4523..83ed544 100644
> --- a/drivers/ata/libahci.c
> +++ b/drivers/ata/libahci.c
> @@ -539,6 +539,27 @@ void ahci_start_engine(struct ata_port *ap)
> {
> void __iomem *port_mmio = ahci_port_base(ap);
> u32 tmp;
> + u8 status;
> +
> + status = readl(port_mmio + PORT_TFDATA)& 0xFF;
> +
> + /*
> + * At end of section 10.1 of AHCI spec (rev 1.3), it states
> + * Software shall not set PxCMD.ST to 1 until it is determined
> + * that a functoinal device is present on the port as determined by
> + * PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h
> + *
> + * Even though most AHCI host controllers work without this check,
> + * specific controller will fail under this condition
> + */
> + if (status& (ATA_BUSY | ATA_DRQ))
> + return;
> + else {
> + ahci_scr_read(&ap->link, SCR_STATUS,&tmp);
> +
> + if ((tmp& 0xf) != 0x3)
> + return;
> + }

Looks good, I'll get this queued up...

2011-04-23 14:40:47

by Tejun Heo

Subject: Re: [PATCH 1/1] libata: ahci_start_engine compliant to AHCI spec

On Sat, Apr 23, 2011 at 8:58 AM, Jian Peng <[email protected]> wrote:
> At the end of section 10.1 of AHCI spec (rev 1.3), it states
>
> Software shall not set PxCMD.ST to 1 until it is determined that
> a functoinal device is present on the port as determined by
> PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h
>
> Even though most AHCI host controller works without this check,
> specific controller will fail under this condition.
>
> Signed-off-by: Jian Peng <[email protected]>

Acked-by: Tejun Heo <[email protected]>

Thanks.

--
tejun

2011-04-25 10:27:58

by Sergei Shtylyov

Subject: Re: [PATCH 1/1] libata: ahci_start_engine compliant to AHCI spec

Hello.

On 23-04-2011 10:58, Jian Peng wrote:

> At the end of section 10.1 of AHCI spec (rev 1.3), it states

> Software shall not set PxCMD.ST to 1 until it is determined that
> a functoinal device is present on the port as determined by
> PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h

> Even though most AHCI host controller works without this check,

"Most" requires plural noun, no?

> specific controller will fail under this condition.

> Signed-off-by: Jian Peng<[email protected]>
[...]

> diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
> index 26d4523..83ed544 100644
> --- a/drivers/ata/libahci.c
> +++ b/drivers/ata/libahci.c
> @@ -539,6 +539,27 @@ void ahci_start_engine(struct ata_port *ap)
> {
> void __iomem *port_mmio = ahci_port_base(ap);
> u32 tmp;
> + u8 status;
> +
> + status = readl(port_mmio + PORT_TFDATA) & 0xFF;

Masking is not necessary.

> +
> + /*
> + * At end of section 10.1 of AHCI spec (rev 1.3), it states
> + * Software shall not set PxCMD.ST to 1 until it is determined
> + * that a functoinal device is present on the port as determined by
> + * PxTFD.STS.BSY=0, PxTFD.STS.DRQ=0 and PxSSTS.DET=3h
> + *
> + * Even though most AHCI host controllers work without this check,

Same about grammar here.

> + * specific controller will fail under this condition
> + */
> + if (status& (ATA_BUSY | ATA_DRQ))
> + return;
> + else {

{} should be used on both branches, according to CodingStyle.

WBR, Sergei