linux-ide.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Tejun Heo <tj@kernel.org>
To: Jeff Garzik <jgarzik@pobox.com>
Cc: Robert Hancock <hancockrwd@gmail.com>,
	Mark Lord <kernel@teksavvy.com>, Seed <seedrubbish@gmail.com>,
	IDE/ATA development list <linux-ide@vger.kernel.org>,
	stable@kernel.org
Subject: [PATCH #upstream-fixes] libahci: fix result_tf handling after an ATA PIO data-in command
Date: Thu, 14 Oct 2010 10:53:35 +0200	[thread overview]
Message-ID: <4CB6C50F.60609@kernel.org> (raw)
In-Reply-To: <AANLkTin=QWXcuy6JOi1y26dWH8FG1ji3JewTXu0K=4ri@mail.gmail.com>

ATA devices don't send D2H Reg FIS after an successful ATA PIO data-in
command.  The host is supposed to take the E_Status of the preceding
PIO Setup FIS.  Update ahci_qc_fill_rtf() such that it takes E_Status
after a successful ATA PIO data-in command.

Without this patch, result_tf for such a command is filled with the
content of the previous D2H Reg FIS which belongs to a previous
command, which can make the command incorrectly seen as failed.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Mark Lord <kernel@teksavvy.com>
Cc: Robert Hancock <hancockrwd@gmail.com>
Cc: stable@kernel.org
---
 drivers/ata/ahci.h    |    1 +
 drivers/ata/libahci.c |   17 ++++++++++++++---
 2 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index e5fdeeb..d1a0f5b 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h
@@ -72,6 +72,7 @@ enum {
 	AHCI_CMD_RESET		= (1 << 8),
 	AHCI_CMD_CLR_BUSY	= (1 << 10),

+	RX_FIS_PIO_SETUP	= 0x20,	/* offset of PIO Setup FIS data */
 	RX_FIS_D2H_REG		= 0x40,	/* offset of D2H Register FIS data */
 	RX_FIS_SDB		= 0x58, /* offset of SDB FIS data */
 	RX_FIS_UNK		= 0x60, /* offset of Unknown FIS data */
diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index 8eea309..0c482a0 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -1830,12 +1830,23 @@ static unsigned int ahci_qc_issue(struct ata_queued_cmd *qc)
 static bool ahci_qc_fill_rtf(struct ata_queued_cmd *qc)
 {
 	struct ahci_port_priv *pp = qc->ap->private_data;
-	u8 *d2h_fis = pp->rx_fis + RX_FIS_D2H_REG;
+	u8 *rx_fis = pp->rx_fis;

 	if (pp->fbs_enabled)
-		d2h_fis += qc->dev->link->pmp * AHCI_RX_FIS_SZ;
+		rx_fis += qc->dev->link->pmp * AHCI_RX_FIS_SZ;
+
+	/*
+	 * After a successful execution of an ATA PIO data-in command,
+	 * the device doesn't send D2H Reg FIS to update the TF and
+	 * the host should take Status from E_Status of the preceding
+	 * PIO Setup FIS.
+	 */
+	if (qc->tf.protocol == ATA_PROT_PIO && qc->dma_dir == DMA_FROM_DEVICE &&
+	    !(qc->flags & ATA_QCFLAG_FAILED))
+		qc->result_tf.command = (rx_fis + RX_FIS_PIO_SETUP)[15];
+	else
+		ata_tf_from_fis(rx_fis + RX_FIS_D2H_REG, &qc->result_tf);

-	ata_tf_from_fis(d2h_fis, &qc->result_tf);
 	return true;
 }


  reply	other threads:[~2010-10-14  8:54 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-09-24  5:34 [BUG] libahci returns stale result tf much of the time Mark Lord
2010-09-24  5:49 ` Seed
2010-09-24  6:27   ` Mark Lord
2010-09-24  7:01     ` Seed
2010-09-24 13:11       ` Mark Lord
2010-09-24 13:24         ` Mark Lord
2010-09-24 23:26           ` Robert Hancock
2010-10-04  9:18             ` Tejun Heo
2010-10-04 17:01               ` Mark Lord
2010-10-04 17:06                 ` Mark Lord
2010-10-04 17:31                   ` Mark Lord
2010-10-04 18:50                     ` Mark Lord
2010-10-04 19:27                       ` Jeff Garzik
2010-10-04 19:35                         ` Mark Lord
2010-10-04 19:42                           ` Jeff Garzik
2010-10-04 19:51                             ` Mark Lord
2010-10-05  7:07                               ` Tejun Heo
2010-10-05 14:06                                 ` Mark Lord
2010-10-05 16:06                                   ` Tejun Heo
2010-10-06  1:00                                     ` Robert Hancock
2010-10-14  8:53                                       ` Tejun Heo [this message]
2010-10-14 21:38                                         ` [PATCH #upstream-fixes] libahci: fix result_tf handling after an ATA PIO data-in command Mark Lord
2010-10-15  0:24                                           ` Robert Hancock
2010-10-15  8:45                                             ` Tejun Heo
2010-10-15 22:14                                               ` Mark Lord
2010-10-15  0:32                                         ` Robert Hancock
2010-10-15  8:45                                           ` Tejun Heo
2010-10-15  9:00                                         ` [PATCH UPDATED " Tejun Heo

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=4CB6C50F.60609@kernel.org \
    --to=tj@kernel.org \
    --cc=hancockrwd@gmail.com \
    --cc=jgarzik@pobox.com \
    --cc=kernel@teksavvy.com \
    --cc=linux-ide@vger.kernel.org \
    --cc=seedrubbish@gmail.com \
    --cc=stable@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).