From: David Bauer Date: Wed, 3 Mar 2021 16:23:16 +0000 (+0100) Subject: ath79: replace chipselect workaround with a proper patch X-Git-Url: http://git.lede-project.org./?a=commitdiff_plain;h=82b9713ce8f5ff90d1d89dbf667aeab5b0aec358;p=openwrt%2Fstaging%2Frobimarko.git ath79: replace chipselect workaround with a proper patch Debugging the SPI CS issue with kernel 5.10 resulted in a better understanding for the root cause and a proper patch with a better explanation. Exchange the old hack patch with a more efficient (and upstreamable) solution. Signed-off-by: David Bauer --- diff --git a/target/linux/ath79/patches-5.10/0054-spi-always-write-cs.patch b/target/linux/ath79/patches-5.10/0054-spi-always-write-cs.patch deleted file mode 100644 index 1e461cee0a..0000000000 --- a/target/linux/ath79/patches-5.10/0054-spi-always-write-cs.patch +++ /dev/null @@ -1,31 +0,0 @@ -Upstream commit d40f0b6f2e21 ("spi: Avoid settingthe chip select if -we don't need to") causes the SPI CS only to be asserted once and not -if it's state stays the same. - -This seems to cause problems with the SPI on the AR724x and the AR913x -(but not the AR71xx). AR934x and subsequent chips do not look affected. - -ToDo: - - Analyze if this is a hardware bug or a bug in the software. - - Send a cleaned up patch upstream. - -Signed-off-by: David Bauer - ---- a/drivers/spi/spi.c -+++ b/drivers/spi/spi.c -@@ -791,6 +791,7 @@ static void spi_set_cs(struct spi_device - { - bool enable1 = enable; - -+#if 0 - /* - * Avoid calling into the driver (or doing delays) if the chip select - * isn't actually changing from the last time this was called. -@@ -801,6 +802,7 @@ static void spi_set_cs(struct spi_device - - spi->controller->last_cs_enable = enable; - spi->controller->last_cs_mode_high = spi->mode & SPI_CS_HIGH; -+#endif - - if (!spi->controller->set_cs_timing) { - if (enable1) diff --git a/target/linux/ath79/patches-5.10/0054-spi-sync-up-initial-chipselect-state.patch b/target/linux/ath79/patches-5.10/0054-spi-sync-up-initial-chipselect-state.patch new file mode 100644 index 0000000000..7329ae5827 --- /dev/null +++ b/target/linux/ath79/patches-5.10/0054-spi-sync-up-initial-chipselect-state.patch @@ -0,0 +1,89 @@ +From 4b7d7f85abac1e7ad9e8b745694e470f0729f527 Mon Sep 17 00:00:00 2001 +From: David Bauer +Date: Wed, 3 Mar 2021 17:11:34 +0100 +Subject: [PATCH] spi: sync up initial chipselect state + +When initially probing the SPI slave device, the call for disabling an +SPI device without the SPI_CS_HIGH flag is not applied, as the +condition for checking whether or not the state to be applied equals the +one currently set evaluates to true. + +This however might not necessarily be the case, as the chipselect might +be active. + +Add a force flag to spi_set_cs which allows to override this +early access condition. Set it to false everywhere except when called +from spi_setup to sync up the initial CS state. + +Fixes commit d40f0b6f2e21 ("spi: Avoid setting the chip select if we don't +need to") + +Signed-off-by: David Bauer +--- + drivers/spi/spi.c | 16 ++++++++-------- + 1 file changed, 8 insertions(+), 8 deletions(-) + +--- a/drivers/spi/spi.c ++++ b/drivers/spi/spi.c +@@ -787,7 +787,7 @@ int spi_register_board_info(struct spi_b + + /*-------------------------------------------------------------------------*/ + +-static void spi_set_cs(struct spi_device *spi, bool enable) ++static void spi_set_cs(struct spi_device *spi, bool enable, bool force) + { + bool enable1 = enable; + +@@ -795,7 +795,7 @@ static void spi_set_cs(struct spi_device + * Avoid calling into the driver (or doing delays) if the chip select + * isn't actually changing from the last time this was called. + */ +- if ((spi->controller->last_cs_enable == enable) && ++ if (!force && (spi->controller->last_cs_enable == enable) && + (spi->controller->last_cs_mode_high == (spi->mode & SPI_CS_HIGH))) + return; + +@@ -1243,7 +1243,7 @@ static int spi_transfer_one_message(stru + struct spi_statistics *statm = &ctlr->statistics; + struct spi_statistics *stats = &msg->spi->statistics; + +- spi_set_cs(msg->spi, true); ++ spi_set_cs(msg->spi, true, false); + + SPI_STATISTICS_INCREMENT_FIELD(statm, messages); + SPI_STATISTICS_INCREMENT_FIELD(stats, messages); +@@ -1311,9 +1311,9 @@ fallback_pio: + &msg->transfers)) { + keep_cs = true; + } else { +- spi_set_cs(msg->spi, false); ++ spi_set_cs(msg->spi, false, false); + _spi_transfer_cs_change_delay(msg, xfer); +- spi_set_cs(msg->spi, true); ++ spi_set_cs(msg->spi, true, false); + } + } + +@@ -1322,7 +1322,7 @@ fallback_pio: + + out: + if (ret != 0 || !keep_cs) +- spi_set_cs(msg->spi, false); ++ spi_set_cs(msg->spi, false, false); + + if (msg->status == -EINPROGRESS) + msg->status = ret; +@@ -3400,11 +3400,11 @@ int spi_setup(struct spi_device *spi) + */ + status = 0; + +- spi_set_cs(spi, false); ++ spi_set_cs(spi, false, true); + pm_runtime_mark_last_busy(spi->controller->dev.parent); + pm_runtime_put_autosuspend(spi->controller->dev.parent); + } else { +- spi_set_cs(spi, false); ++ spi_set_cs(spi, false, true); + } + + mutex_unlock(&spi->controller->io_mutex);