* [PATCH] spi: fix cs_change for last transfer
@ 2020-03-15 17:08 Michał Mirosław
2020-03-16 12:17 ` Mark Brown
0 siblings, 1 reply; 4+ messages in thread
From: Michał Mirosław @ 2020-03-15 17:08 UTC (permalink / raw)
To: Mark Brown; +Cc: linux-spi, linux-kernel
Generic spi_transfer_one_message() implementation introduced in
commit b158935f70b9 has a bug in cs_change handling: it keeps CS
asserted when cs_change is set. Fix it.
Cc: stable@vger.kernel.org
Fixes: b158935f70b9 ("spi: Provide common spi_message processing loop")
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
---
drivers/spi/spi.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index 8994545367a2..5012eabde468 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -1206,7 +1206,7 @@ static int spi_transfer_one_message(struct spi_controller *ctlr,
struct spi_message *msg)
{
struct spi_transfer *xfer;
- bool keep_cs = false;
+ bool keep_cs = true;
int ret = 0;
struct spi_statistics *statm = &ctlr->statistics;
struct spi_statistics *stats = &msg->spi->statistics;
@@ -1268,7 +1268,7 @@ static int spi_transfer_one_message(struct spi_controller *ctlr,
if (xfer->cs_change) {
if (list_is_last(&xfer->transfer_list,
&msg->transfers)) {
- keep_cs = true;
+ keep_cs = false;
} else {
spi_set_cs(msg->spi, false);
_spi_transfer_cs_change_delay(msg, xfer);
--
2.20.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH] spi: fix cs_change for last transfer
2020-03-15 17:08 [PATCH] spi: fix cs_change for last transfer Michał Mirosław
@ 2020-03-16 12:17 ` Mark Brown
2020-03-16 14:34 ` Michał Mirosław
0 siblings, 1 reply; 4+ messages in thread
From: Mark Brown @ 2020-03-16 12:17 UTC (permalink / raw)
To: Michał Mirosław; +Cc: linux-spi, linux-kernel
[-- Attachment #1: Type: text/plain, Size: 965 bytes --]
On Sun, Mar 15, 2020 at 06:08:53PM +0100, Michał Mirosław wrote:
> Generic spi_transfer_one_message() implementation introduced in
> commit b158935f70b9 has a bug in cs_change handling: it keeps CS
> asserted when cs_change is set. Fix it.
> struct spi_transfer *xfer;
> - bool keep_cs = false;
> + bool keep_cs = true;
> int ret = 0;
> struct spi_statistics *statm = &ctlr->statistics;
> struct spi_statistics *stats = &msg->spi->statistics;
> @@ -1268,7 +1268,7 @@ static int spi_transfer_one_message(struct spi_controller *ctlr,
> if (xfer->cs_change) {
> if (list_is_last(&xfer->transfer_list,
> &msg->transfers)) {
> - keep_cs = true;
> + keep_cs = false;
> } else {
This is the opposite of the intended behaviour - we want to deassert
chip select at the end of the message unless cs_change is set on the
last transfer. If this were broken I would expect to see widespread
problems being reported.
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] spi: fix cs_change for last transfer
2020-03-16 12:17 ` Mark Brown
@ 2020-03-16 14:34 ` Michał Mirosław
2020-03-16 15:13 ` Mark Brown
0 siblings, 1 reply; 4+ messages in thread
From: Michał Mirosław @ 2020-03-16 14:34 UTC (permalink / raw)
To: Mark Brown; +Cc: linux-spi, linux-kernel
On Mon, Mar 16, 2020 at 12:17:50PM +0000, Mark Brown wrote:
> On Sun, Mar 15, 2020 at 06:08:53PM +0100, Michał Mirosław wrote:
> > Generic spi_transfer_one_message() implementation introduced in
> > commit b158935f70b9 has a bug in cs_change handling: it keeps CS
> > asserted when cs_change is set. Fix it.
>
> > struct spi_transfer *xfer;
> > - bool keep_cs = false;
> > + bool keep_cs = true;
> > int ret = 0;
> > struct spi_statistics *statm = &ctlr->statistics;
> > struct spi_statistics *stats = &msg->spi->statistics;
> > @@ -1268,7 +1268,7 @@ static int spi_transfer_one_message(struct spi_controller *ctlr,
> > if (xfer->cs_change) {
> > if (list_is_last(&xfer->transfer_list,
> > &msg->transfers)) {
> > - keep_cs = true;
> > + keep_cs = false;
> > } else {
>
> This is the opposite of the intended behaviour - we want to deassert
> chip select at the end of the message unless cs_change is set on the
> last transfer. If this were broken I would expect to see widespread
> problems being reported.
This is unfortunate naming I suppose. I reread the spi.h comments
a few more times and it seems indeed, that .cs_change == 1 on last
transfer means to a driver: "you may leave CS unchanged" - quite the
reverse compared to non-last transfers.
Please drop this patch then.
Best Regards
Michał Mirosław
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] spi: fix cs_change for last transfer
2020-03-16 14:34 ` Michał Mirosław
@ 2020-03-16 15:13 ` Mark Brown
0 siblings, 0 replies; 4+ messages in thread
From: Mark Brown @ 2020-03-16 15:13 UTC (permalink / raw)
To: Michał Mirosław; +Cc: linux-spi, linux-kernel
[-- Attachment #1: Type: text/plain, Size: 792 bytes --]
On Mon, Mar 16, 2020 at 03:34:55PM +0100, Michał Mirosław wrote:
> On Mon, Mar 16, 2020 at 12:17:50PM +0000, Mark Brown wrote:
> > This is the opposite of the intended behaviour - we want to deassert
> > chip select at the end of the message unless cs_change is set on the
> > last transfer. If this were broken I would expect to see widespread
> > problems being reported.
> This is unfortunate naming I suppose. I reread the spi.h comments
> a few more times and it seems indeed, that .cs_change == 1 on last
> transfer means to a driver: "you may leave CS unchanged" - quite the
> reverse compared to non-last transfers.
cs_change also means that we should add an extra chip select transition
on transfers other than the last.
> Please drop this patch then.
OK.
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2020-03-16 15:13 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-03-15 17:08 [PATCH] spi: fix cs_change for last transfer Michał Mirosław
2020-03-16 12:17 ` Mark Brown
2020-03-16 14:34 ` Michał Mirosław
2020-03-16 15:13 ` Mark Brown
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).