When 7c7f9bc986e6 ("serial: Deassert Transmit Enable on probe in driver-specific way") got backported to 5.15.y, there known as b079d3775237, this hunk was accidentally left out. So if the "goto failed_get_rs485;" is hit, the cleanup will do uart_remove_one_port() despite uart_add_one_port() not having been called.
Add the missing hunk.
Fixes: b079d3775237 ("serial: Deassert Transmit Enable on probe in driver-specific way") Signed-off-by: Rasmus Villemoes linux@rasmusvillemoes.dk ---
Not quite sure how to submit patches for a specific -stable series only, or if the Fixes tag is appropriate and correct. Please let me know if you'd have preferred anything different.
drivers/tty/serial/fsl_lpuart.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/tty/serial/fsl_lpuart.c b/drivers/tty/serial/fsl_lpuart.c index 595430aedc0d..fc311df9f1c9 100644 --- a/drivers/tty/serial/fsl_lpuart.c +++ b/drivers/tty/serial/fsl_lpuart.c @@ -2784,9 +2784,9 @@ static int lpuart_probe(struct platform_device *pdev) return 0;
failed_irq_request: -failed_get_rs485: uart_remove_one_port(&lpuart_reg, &sport->port); failed_attach_port: +failed_get_rs485: failed_reset: lpuart_disable_clks(sport); return ret;
base-commit: fd6d66840b4269da4e90e1ea807ae3197433bc66
When 7c7f9bc986e6 ("serial: Deassert Transmit Enable on probe in driver-specific way") got backported to 5.15.y, there known as b079d3775237, some hunks were accidentally left out.
In fsl_lpuart.c, this amounts to uart_remove_one_port() being called in an error path despite uart_add_one_port() not having been called.
In serial_core.c, it is possible that the omission in uart_suspend_port() is harmless, but the backport did have the corresponding hunk in uart_resume_port(), it runs counter to the original commit's intention of
Skip any invocation of ->set_mctrl() if RS485 is enabled.
and it's certainly better to be aligned with upstream.
Fixes: b079d3775237 ("serial: Deassert Transmit Enable on probe in driver-specific way") Signed-off-by: Rasmus Villemoes linux@rasmusvillemoes.dk ---
v2: Also amend uart_suspend_port(), update commit log accordingly.
drivers/tty/serial/fsl_lpuart.c | 2 +- drivers/tty/serial/serial_core.c | 3 ++- 2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/tty/serial/fsl_lpuart.c b/drivers/tty/serial/fsl_lpuart.c index 595430aedc0d..fc311df9f1c9 100644 --- a/drivers/tty/serial/fsl_lpuart.c +++ b/drivers/tty/serial/fsl_lpuart.c @@ -2784,9 +2784,9 @@ static int lpuart_probe(struct platform_device *pdev) return 0;
failed_irq_request: -failed_get_rs485: uart_remove_one_port(&lpuart_reg, &sport->port); failed_attach_port: +failed_get_rs485: failed_reset: lpuart_disable_clks(sport); return ret; diff --git a/drivers/tty/serial/serial_core.c b/drivers/tty/serial/serial_core.c index 5f8f0a90ce55..45b721abaa2f 100644 --- a/drivers/tty/serial/serial_core.c +++ b/drivers/tty/serial/serial_core.c @@ -2225,7 +2225,8 @@ int uart_suspend_port(struct uart_driver *drv, struct uart_port *uport)
spin_lock_irq(&uport->lock); ops->stop_tx(uport); - ops->set_mctrl(uport, 0); + if (!(uport->rs485.flags & SER_RS485_ENABLED)) + ops->set_mctrl(uport, 0); ops->stop_rx(uport); spin_unlock_irq(&uport->lock);
base-commit: 5827ddaf4534c52d31dd464679a186b41810ef76
Rasmus Villemoes wrote on Thu, Dec 22, 2022 at 12:44:14PM +0100:
When 7c7f9bc986e6 ("serial: Deassert Transmit Enable on probe in driver-specific way") got backported to 5.15.y, there known as b079d3775237, some hunks were accidentally left out.
In fsl_lpuart.c, this amounts to uart_remove_one_port() being called in an error path despite uart_add_one_port() not having been called.
In serial_core.c, it is possible that the omission in uart_suspend_port() is harmless, but the backport did have the corresponding hunk in uart_resume_port(), it runs counter to the original commit's intention of
Skip any invocation of ->set_mctrl() if RS485 is enabled.
and it's certainly better to be aligned with upstream.
Fixes: b079d3775237 ("serial: Deassert Transmit Enable on probe in driver-specific way") Signed-off-by: Rasmus Villemoes linux@rasmusvillemoes.dk
Thank you! I've confirmed both hunks are part of the original patch; I'm not quite sure why we missed them...
Reviewed-by: Dominique MARTINET dominique.martinet@atmark-techno.com
This doesn't apply cleanly to 5.10 because 5.10 did not get 401fb66a3 ("fsl_lpuart: Don't enable interrupts too early") backported, I think it makes sense to take as well so I'll send a backport of these two patches for 5.10
On Thu, Dec 22, 2022 at 12:44:14PM +0100, Rasmus Villemoes wrote:
When 7c7f9bc986e6 ("serial: Deassert Transmit Enable on probe in driver-specific way") got backported to 5.15.y, there known as b079d3775237, some hunks were accidentally left out.
In fsl_lpuart.c, this amounts to uart_remove_one_port() being called in an error path despite uart_add_one_port() not having been called.
In serial_core.c, it is possible that the omission in uart_suspend_port() is harmless, but the backport did have the corresponding hunk in uart_resume_port(), it runs counter to the original commit's intention of
Skip any invocation of ->set_mctrl() if RS485 is enabled.
and it's certainly better to be aligned with upstream.
Fixes: b079d3775237 ("serial: Deassert Transmit Enable on probe in driver-specific way") Signed-off-by: Rasmus Villemoes linux@rasmusvillemoes.dk
v2: Also amend uart_suspend_port(), update commit log accordingly.
Now queued up, thanks.
greg k-h
linux-stable-mirror@lists.linaro.org
-
Dominique Martinet -
Greg Kroah-Hartman -
Rasmus Villemoes