All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg KH <gregkh@suse.de>
To: linux-kernel@vger.kernel.org, stable@kernel.org, jejb@kernel.org
Cc: Justin Forbes <jmforbes@linuxtx.org>,
	Zwane Mwaikambo <zwane@arm.linux.org.uk>,
	"Theodore Ts'o" <tytso@mit.edu>,
	Randy Dunlap <rdunlap@xenotime.net>,
	Dave Jones <davej@redhat.com>,
	Chuck Wolber <chuckw@quantumlinux.com>,
	Chris Wedgwood <reviews@ml.cw.f00f.org>,
	Michael Krufky <mkrufky@linuxtv.org>,
	Chuck Ebbert <cebbert@redhat.com>,
	Domenico Andreoli <cavokz@gmail.com>, Willy Tarreau <w@1wt.eu>,
	Rodrigo Rubira Branco <rbranco@la.checkpoint.com>,
	Jake Edge <jake@lwn.net>, Eugene Teo <eteo@redhat.com>,
	torvalds@linux-foundation.org, akpm@linux-foundation.org,
	alan@lxorguk.ukuu.org.uk, Ned Forrester <nforrester@whoi.edu>,
	Vernon Sauder <vernoninhand@gmail.com>,
	Eric Miao <eric.y.miao@gmail.com>,
	David Brownell <dbrownell@users.sourceforge.net>
Subject: [patch 03/28] pxa2xx_spi: chipselect bugfixes
Date: Mon, 6 Oct 2008 16:17:03 -0700	[thread overview]
Message-ID: <20081006231703.GD20567@suse.de> (raw)
In-Reply-To: <20081006231639.GA20567@suse.de>

[-- Attachment #1: pxa2xx_spi-chipselect-bugfixes.patch --]
[-- Type: text/plain, Size: 5814 bytes --]

2.6.25-stable review patch.  If anyone has any objections, please let us
know.

------------------
From: Ned Forrester <nforrester@whoi.edu>

commit 8423597d676615f3dd2d9ab36f59f147086b90b8 upstream

Fixes several chipselect bugs in the pxa2xx_spi driver.  These bugs are in
all versions of this driver and prevent using it with chips like m25p16
flash.

 1. The spi_transfer.cs_change flag is handled too early:
    before spi_transfer.delay_usecs applies, thus making the
    delay ineffective at holding chip select.

 2. spi_transfer.delay_usecs is ignored on the last transfer
    of a message (likewise not holding chipselect long enough).

 3. If spi_transfer.cs_change is set on the last transfer, the
    chip select is always disabled, instead of the intended
    meaning: optionally holding chip select enabled for the
    next message.

Those first three bugs were fixed with a relocation of delays
and chip select de-assertions.

 4. If a message has the cs_change flag set on the last transfer,
    and had the chip select stayed enabled as requested (see 3,
    above), it would not have been disabled if the next message is
    for a different chip.  Fixed by dropping chip select regardless
    of cs_change at end of a message, if there is no next message
    or if the next message is for a different chip.

This patch should apply to all kernels back to and including 2.6.20;
it was test patched against 2.6.20.  An additional patch would be
required for older kernels, but those versions are very buggy anyway.

Signed-off-by: Ned Forrester <nforrester@whoi.edu>
Cc: Vernon Sauder <vernoninhand@gmail.com>
Cc: Eric Miao <eric.y.miao@gmail.com>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

---
 drivers/spi/pxa2xx_spi.c |   59 ++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 48 insertions(+), 11 deletions(-)

--- a/drivers/spi/pxa2xx_spi.c
+++ b/drivers/spi/pxa2xx_spi.c
@@ -143,7 +143,6 @@ struct driver_data {
 	size_t tx_map_len;
 	u8 n_bytes;
 	u32 dma_width;
-	int cs_change;
 	int (*write)(struct driver_data *drv_data);
 	int (*read)(struct driver_data *drv_data);
 	irqreturn_t (*transfer_handler)(struct driver_data *drv_data);
@@ -405,8 +404,45 @@ static void giveback(struct driver_data 
 					struct spi_transfer,
 					transfer_list);
 
+	/* Delay if requested before any change in chip select */
+	if (last_transfer->delay_usecs)
+		udelay(last_transfer->delay_usecs);
+
+	/* Drop chip select UNLESS cs_change is true or we are returning
+	 * a message with an error, or next message is for another chip
+	 */
 	if (!last_transfer->cs_change)
 		drv_data->cs_control(PXA2XX_CS_DEASSERT);
+	else {
+		struct spi_message *next_msg;
+
+		/* Holding of cs was hinted, but we need to make sure
+		 * the next message is for the same chip.  Don't waste
+		 * time with the following tests unless this was hinted.
+		 *
+		 * We cannot postpone this until pump_messages, because
+		 * after calling msg->complete (below) the driver that
+		 * sent the current message could be unloaded, which
+		 * could invalidate the cs_control() callback...
+		 */
+
+		/* get a pointer to the next message, if any */
+		spin_lock_irqsave(&drv_data->lock, flags);
+		if (list_empty(&drv_data->queue))
+			next_msg = NULL;
+		else
+			next_msg = list_entry(drv_data->queue.next,
+					struct spi_message, queue);
+		spin_unlock_irqrestore(&drv_data->lock, flags);
+
+		/* see if the next and current messages point
+		 * to the same chip
+		 */
+		if (next_msg && next_msg->spi != msg->spi)
+			next_msg = NULL;
+		if (!next_msg || msg->state == ERROR_STATE)
+			drv_data->cs_control(PXA2XX_CS_DEASSERT);
+	}
 
 	msg->state = NULL;
 	if (msg->complete)
@@ -489,10 +525,9 @@ static void dma_transfer_complete(struct
 	msg->actual_length += drv_data->len -
 				(drv_data->rx_end - drv_data->rx);
 
-	/* Release chip select if requested, transfer delays are
-	 * handled in pump_transfers */
-	if (drv_data->cs_change)
-		drv_data->cs_control(PXA2XX_CS_DEASSERT);
+	/* Transfer delays and chip select release are
+	 * handled in pump_transfers or giveback
+	 */
 
 	/* Move to next transfer */
 	msg->state = next_transfer(drv_data);
@@ -601,10 +636,9 @@ static void int_transfer_complete(struct
 	drv_data->cur_msg->actual_length += drv_data->len -
 				(drv_data->rx_end - drv_data->rx);
 
-	/* Release chip select if requested, transfer delays are
-	 * handled in pump_transfers */
-	if (drv_data->cs_change)
-		drv_data->cs_control(PXA2XX_CS_DEASSERT);
+	/* Transfer delays and chip select release are
+	 * handled in pump_transfers or giveback
+	 */
 
 	/* Move to next transfer */
 	drv_data->cur_msg->state = next_transfer(drv_data);
@@ -838,13 +872,17 @@ static void pump_transfers(unsigned long
 		return;
 	}
 
-	/* Delay if requested at end of transfer*/
+	/* Delay if requested at end of transfer before CS change */
 	if (message->state == RUNNING_STATE) {
 		previous = list_entry(transfer->transfer_list.prev,
 					struct spi_transfer,
 					transfer_list);
 		if (previous->delay_usecs)
 			udelay(previous->delay_usecs);
+
+		/* Drop chip select only if cs_change is requested */
+		if (previous->cs_change)
+			drv_data->cs_control(PXA2XX_CS_DEASSERT);
 	}
 
 	/* Check for transfers that need multiple DMA segments */
@@ -889,7 +927,6 @@ static void pump_transfers(unsigned long
 	drv_data->len = transfer->len & DCMD_LENGTH;
 	drv_data->write = drv_data->tx ? chip->write : null_writer;
 	drv_data->read = drv_data->rx ? chip->read : null_reader;
-	drv_data->cs_change = transfer->cs_change;
 
 	/* Change speed and bit per word on a per transfer */
 	cr0 = chip->cr0;

-- 

  parent reply	other threads:[~2008-10-06 23:22 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20081006225949.357365087@mini.kroah.org>
2008-10-06 23:16 ` [patch 00/28] 2.6.25-stable review Greg KH
2008-10-06 23:16   ` [patch 01/28] USB: fix hcd interrupt disabling Greg KH
2008-10-06 23:17   ` [patch 02/28] pxa2xx_spi: dma bugfixes Greg KH
2008-10-07  0:15     ` Ned Forrester
2008-10-06 23:17   ` Greg KH [this message]
2008-10-06 23:17   ` [patch 04/28] drivers/mmc/card/block.c: fix refcount leak in mmc_block_open() Greg KH
2008-10-06 23:17   ` [patch 05/28] ALSA: hda - Fix model for Dell Inspiron 1525 Greg KH
2008-10-07  5:55     ` Takashi Iwai
2008-10-09  2:38       ` Greg KH
2008-10-06 23:17   ` [patch 06/28] i2c-dev: Return correct error code on class_create() failure Greg KH
2008-10-06 23:17   ` [patch 07/28] ACPI: Fix thermal shutdowns Greg KH
2008-10-06 23:17   ` [patch 08/28] x86: add io delay quirk for Presario F700 Greg KH
2008-10-06 23:17   ` [patch 09/28] rtc: fix deadlock Greg KH
2008-10-06 23:17   ` [patch 10/28] ACPI: Avoid bogus EC timeout when EC is in Polling mode Greg KH
2008-10-06 23:17   ` [patch 11/28] clockevents: prevent clockevent event_handler ending up handler_noop Greg KH
2008-10-06 23:17   ` [patch 12/28] clockevents: prevent endless loop in periodic broadcast handler Greg KH
2008-10-06 23:17   ` [patch 13/28] clockevents: enforce reprogram in oneshot setup Greg KH
2008-10-06 23:17   ` [patch 14/28] clockevents: prevent multiple init/shutdown Greg KH
2008-10-06 23:17   ` [patch 15/28] clockevents: prevent endless loop lockup Greg KH
2008-10-06 23:17   ` [patch 16/28] HPET: make minimum reprogramming delta useful Greg KH
2008-10-06 23:17   ` [patch 17/28] clockevents: broadcast fixup possible waiters Greg KH
2008-10-06 23:17   ` [patch 18/28] x86: HPET fix moronic 32/64bit thinko Greg KH
2008-10-06 23:17   ` [patch 19/28] x86: HPET: read back compare register before reading counter Greg KH
2008-10-06 23:17   ` [patch 20/28] ntp: fix calculation of the next jiffie to trigger RTC sync Greg KH
2008-10-06 23:17   ` [patch 21/28] clockevents: remove WARN_ON which was used to gather information Greg KH
2008-10-06 23:17   ` [patch 22/28] x86: Fix broken LDT access in VMI Greg KH
2008-10-06 23:17   ` [patch 23/28] ipv6: Fix OOPS in ip6_dst_lookup_tail() Greg KH
2008-10-06 23:51     ` Neil Horman
2008-10-06 23:17   ` [patch 24/28] niu: panic on reset Greg KH
2008-10-06 23:17   ` [patch 25/28] netlink: fix overrun in attribute iteration Greg KH
2008-10-06 23:17   ` [patch 26/28] sctp: do not enable peer features if we cant do them Greg KH
2008-10-06 23:17   ` [patch 27/28] sctp: Fix oops when INIT-ACK indicates that peer doesnt support AUTH Greg KH
2008-10-06 23:18   ` [patch 28/28] udp: Fix rcv socket locking Greg KH

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=20081006231703.GD20567@suse.de \
    --to=gregkh@suse.de \
    --cc=akpm@linux-foundation.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=cavokz@gmail.com \
    --cc=cebbert@redhat.com \
    --cc=chuckw@quantumlinux.com \
    --cc=davej@redhat.com \
    --cc=dbrownell@users.sourceforge.net \
    --cc=eric.y.miao@gmail.com \
    --cc=eteo@redhat.com \
    --cc=jake@lwn.net \
    --cc=jejb@kernel.org \
    --cc=jmforbes@linuxtx.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mkrufky@linuxtv.org \
    --cc=nforrester@whoi.edu \
    --cc=rbranco@la.checkpoint.com \
    --cc=rdunlap@xenotime.net \
    --cc=reviews@ml.cw.f00f.org \
    --cc=stable@kernel.org \
    --cc=torvalds@linux-foundation.org \
    --cc=tytso@mit.edu \
    --cc=vernoninhand@gmail.com \
    --cc=w@1wt.eu \
    --cc=zwane@arm.linux.org.uk \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.