linux.git/drivers/tty, branch v6.6.39

serial: imx: Raise TX trigger level to 8

2024-07-11T10:49:09+00:00

[ Upstream commit a3d8728ab079951741efa11360df43dbfacba7ab ]

At the default TX trigger level of 2 in non-DMA mode (meaning that an
interrupt is generated when less than 2 characters are left in the
FIFO), we have observed frequent buffer underruns at 115200 Baud on an
i.MX8M Nano. This can cause communication issues if the receiving side
expects a continuous transfer.

Increasing the level to 8 makes the UART trigger an interrupt earlier,
giving the kernel enough time to refill the FIFO, at the cost of
triggering one interrupt per ~24 instead of ~30 bytes of transmitted
data (as the i.MX UART has a 32 byte FIFO).

Signed-off-by: Michael Krummsdorf 
Signed-off-by: Matthias Schiffer 
Link: https://lore.kernel.org/r/20240508133744.35858-1-matthias.schiffer@ew.tq-group.com
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Sasha Levin

serial: 8250_omap: Fix Errata i2310 with RX FIFO level check

2024-07-05T07:34:07+00:00

commit c128a1b0523b685c8856ddc0ac0e1caef1fdeee5 upstream.

Errata i2310[0] says, Erroneous timeout can be triggered,
if this Erroneous interrupt is not cleared then it may leads
to storm of interrupts.

Commit 9d141c1e6157 ("serial: 8250_omap: Implementation of Errata i2310")
which added the workaround but missed ensuring RX FIFO is really empty
before applying the errata workaround as recommended in the errata text.
Fix this by adding back check for UART_OMAP_RX_LVL to be 0 for
workaround to take effect.

[0] https://www.ti.com/lit/pdf/sprz536 page 23

Fixes: 9d141c1e6157 ("serial: 8250_omap: Implementation of Errata i2310")
Cc: stable@vger.kernel.org
Reported-by: Vignesh Raghavendra 
Closes: https://lore.kernel.org/all/e96d0c55-0b12-4cbf-9d23-48963543de49@ti.com/
Signed-off-by: Udit Kumar 
Link: https://lore.kernel.org/r/20240625160725.2102194-1-u-kumar1@ti.com
Signed-off-by: Greg Kroah-Hartman

serial: imx: only set receiver level if it is zero

2024-07-05T07:34:07+00:00

commit 9706fc87b4cff0ac4f5d5d62327be83fe72e3108 upstream.

With commit a81dbd0463ec ("serial: imx: set receiver level before
starting uart") we set the receiver level to its default value. This
caused a regression when using SDMA, where the receiver level is 9
instead of 8 (default). This change will first check if the receiver
level is zero and only then set it to the default. This still avoids the
interrupt storm when the receiver level is zero.

Fixes: a81dbd0463ec ("serial: imx: set receiver level before starting uart")
Cc: stable 
Signed-off-by: Stefan Eichenberger 
Link: https://lore.kernel.org/r/20240703112543.148304-1-eichest@gmail.com
Signed-off-by: Greg Kroah-Hartman

tty: mcf: MCF54418 has 10 UARTS

2024-07-05T07:34:00+00:00

commit 7c92a8bd53f24d50c8cf4aba53bb75505b382fed upstream.

Most of the colfires have up to 5 UARTs but MCF54418 has up-to 10 !
Change the maximum value authorized.

Signed-off-by: Jean-Michel Hautbois 
Cc: stable 
Fixes: 2545cf6e94b4 ("m68knommu: allow 4 coldfire serial ports")
Link: https://lore.kernel.org/r/20240620-upstream-uart-v1-1-a9d0d95fb19e@yoseli.org
Signed-off-by: Greg Kroah-Hartman

serial: bcm63xx-uart: fix tx after conversion to uart_port_tx_limited()

2024-07-05T07:34:00+00:00

commit ea55c65dedf40e9c1911dc1e63e26bc9a59692b9 upstream.

When bcm63xx-uart was converted to uart_port_tx_limited(), it implicitly
added a call to stop_tx(). This causes garbage to be put out on the
serial console. To fix this, pass UART_TX_NOSTOP in flags, and manually
call stop_tx() ourselves analogue to how a similar issue was fixed in
commit 7be50f2e8f20 ("serial: mxs-auart: fix tx").

Fixes: d11cc8c3c4b6 ("tty: serial: use uart_port_tx_limited()")
Cc: stable@vger.kernel.org
Signed-off-by: Jonas Gorski 
Signed-off-by: Doug Brown 
Link: https://lore.kernel.org/r/20240606195632.173255-4-doug@schmorgal.com
Signed-off-by: Greg Kroah-Hartman

serial: imx: set receiver level before starting uart

2024-07-05T07:34:00+00:00

commit a81dbd0463eca317eee44985a66aa6cc2ce5c101 upstream.

Set the receiver level to something > 0 before calling imx_uart_start_rx
in rs485_config. This is necessary to avoid an interrupt storm that
might prevent the system from booting. This was seen on an i.MX7 device
when the rs485-rts-active-low property was active in the device tree.

Fixes: 6d215f83e5fc ("serial: imx: warn user when using unsupported configuration")
Cc: stable 
Signed-off-by: Stefan Eichenberger 
Link: https://lore.kernel.org/r/20240621153829.183780-1-eichest@gmail.com
Signed-off-by: Greg Kroah-Hartman

serial: 8250_omap: Implementation of Errata i2310

2024-07-05T07:33:59+00:00

commit 9d141c1e615795eeb93cd35501ad144ee997a826 upstream.

As per Errata i2310[0], Erroneous timeout can be triggered,
if this Erroneous interrupt is not cleared then it may leads
to storm of interrupts, therefore apply Errata i2310 solution.

[0] https://www.ti.com/lit/pdf/sprz536 page 23

Fixes: b67e830d38fa ("serial: 8250: 8250_omap: Fix possible interrupt storm on K3 SoCs")
Cc: stable@vger.kernel.org
Signed-off-by: Udit Kumar 
Link: https://lore.kernel.org/r/20240619105903.165434-1-u-kumar1@ti.com
Signed-off-by: Greg Kroah-Hartman

serial: 8250_dw: Revert "Move definitions to the shared header"

2024-06-27T11:49:13+00:00

commit 2c94512055f362dd789e0f87b8566feeddec83c9 upstream.

This reverts commit d9666dfb314e1ffd6eb9c3c4243fe3e094c047a7.

The container of the struct dw8250_port_data is private to the actual
driver. In particular, 8250_lpss and 8250_dw use different data types
that are assigned to the UART port private_data. Hence, it must not
be used outside the specific driver.

Fix the mistake made in the past by moving the respective definitions
to the specific driver.

Signed-off-by: Andy Shevchenko 
Link: https://lore.kernel.org/r/20240514190730.2787071-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Greg Kroah-Hartman

tty: add the option to have a tty reject a new ldisc

2024-06-27T11:49:05+00:00

[ Upstream commit 6bd23e0c2bb6c65d4f5754d1456bc9a4427fc59b ]

... and use it to limit the virtual terminals to just N_TTY.  They are
kind of special, and in particular, the "con_write()" routine violates
the "writes cannot sleep" rule that some ldiscs rely on.

This avoids the

   BUG: sleeping function called from invalid context at kernel/printk/printk.c:2659

when N_GSM has been attached to a virtual console, and gsmld_write()
calls con_write() while holding a spinlock, and con_write() then tries
to get the console lock.

Tested-by: Tetsuo Handa 
Cc: Jiri Slaby 
Cc: Andrew Morton 
Cc: Daniel Starke 
Reported-by: syzbot 
Closes: https://syzkaller.appspot.com/bug?extid=dbac96d8e73b61aa559c
Signed-off-by: Linus Torvalds 
Link: https://lore.kernel.org/r/20240423163339.59780-1-torvalds@linux-foundation.org
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Sasha Levin

serial: exar: adding missing CTI and Exar PCI ids

2024-06-27T11:49:04+00:00

[ Upstream commit b86ae40ffcf5a16b9569b1016da4a08c4f352ca2 ]

- Added Connect Tech and Exar IDs not already in pci_ids.h

Signed-off-by: Parker Newman 
Link: https://lore.kernel.org/r/7c3d8e795a864dd9b0a00353b722060dc27c4e09.1713270624.git.pnewman@connecttech.com
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Sasha Levin