- 25 Jun, 2016 40 commits
-
-
Krzysztof Kozlowski authored
When the clk_get() of "uart" clock returns EPROBE_DEFER, the next re-probe finishes with success but uses invalid (ERR_PTR) values. This leads to dereferencing of ERR_PTR stored under ourport->clk: 12c30000.serial: Controller clock not found (...) 12c30000.serial: ttySAC3 at MMIO 0x12c30000 (irq = 61, base_baud = 0) is a S3C6400/10 Unable to handle kernel paging request at virtual address fffffdfb (clk_prepare) from [<c039f7d0>] (s3c24xx_serial_pm+0x20/0x128) (s3c24xx_serial_pm) from [<c0395414>] (uart_change_pm+0x38/0x40) (uart_change_pm) from [<c039689c>] (uart_add_one_port+0x31c/0x44c) (uart_add_one_port) from [<c03a035c>] (s3c24xx_serial_probe+0x2a8/0x418) (s3c24xx_serial_probe) from [<c03ee110>] (platform_drv_probe+0x50/0xb0) (platform_drv_probe) from [<c03ecb44>] (driver_probe_device+0x1f4/0x2b0) (driver_probe_device) from [<c03eb0c0>] (bus_for_each_drv+0x44/0x8c) (bus_for_each_drv) from [<c03ec8c8>] (__device_attach+0x9c/0x100) (__device_attach) from [<c03ebf54>] (bus_probe_device+0x84/0x8c) (bus_probe_device) from [<c03ec388>] (deferred_probe_work_func+0x60/0x8c) (deferred_probe_work_func) from [<c012fee4>] (process_one_work+0x120/0x328) (process_one_work) from [<c0130150>] (worker_thread+0x2c/0x4ac) (worker_thread) from [<c0135320>] (kthread+0xd8/0xf4) (kthread) from [<c0107978>] (ret_from_fork+0x14/0x3c) The first unsuccessful clk_get() causes s3c24xx_serial_init_port() to exit with failure but the s3c24xx_uart_port is left half-configured (e.g. port->mapbase is set, clk contains ERR_PTR). On next re-probe, the function s3c24xx_serial_init_port() will exit early with success because of configured port->mapbase and driver will use old values, including the ERR_PTR as clock. Fix this by cleaning the port->mapbase on error path so each re-probe will initialize all of the port settings. Fixes: 60e93575 ("serial: samsung: enable clock before clearing pending interrupts during init") Cc: <stable@vger.kernel.org> Signed-off-by:
Krzysztof Kozlowski <k.kozlowski@samsung.com> Reviewed-by:
Javier Martinez Canillas <javier@osg.samsung.com> Tested-by:
Kevin Hilman <khilman@baylibre.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Ji-Ze Hong (Peter Hong) authored
Some BIOS only use _OSI("Linux") to distinguish between Linux & Windows. Apply Level/Low to UART trigger mode if Windows, Edge/High mode otherwise. But since 2.6.23 the mainline kernel no longer returns true for _OSI(“Linux”). The default IRQ0~15 trigger mode in Linux is Edge/High mode without ACPI MADT override. It mismatches IRQ mode and makes UART malfunctional on such motherboard. This patch will check the current IRQ mode and apply correct mode to UART. The following link is F81216AD spec PDF: http://html.alldatasheet.com/html-pdf/257956/FINTEK/F81216AD/5569/ 25/F81216AD.html LDN0~3 70h: IRQ channel & Mode register Bit 6~5 : 00 : Active low level mode 01 : Active high edge mode Bit 4 : Sharing Flag (0: not share/1: share) Bit 3~0 : IRQ channel Signed-off-by:Ji-Ze Hong (Peter Hong) <hpeter+linux_kernel@gmail.com> Signed-off-by:
-
Geert Uytterhoeven authored
Submitters of device tree binding documentation may forget to CC the subsystem maintainer if this is missing. Signed-off-by:
Geert Uytterhoeven <geert@linux-m68k.org> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: linux-serial@vger.kernel.org Signed-off-by:
-
Purna Chandra Mandal authored
Allocated memory for 'sport->irq_fault_name' is freed twice, first in error check of 'if(!sport->irq_rx_name)' and other in fallback handler. Signed-off-by:
Purna Chandra Mandal <purna.mandal@microchip.com> Signed-off-by:
-
Shubhrajyoti Datta authored
Currently the clocks are enabled at probe and disabled at remove. Instead enable the clocks when used. Signed-off-by:
Shubhrajyoti Datta <shubhraj@xilinx.com> Signed-off-by:
-
Alexandre Belloni authored
When using DMA, half duplex doesn't work properly because rx is not stopped before starting tx. Ensure we call atmel_stop_rx() in the DMA case. Signed-off-by:
Alexandre Belloni <alexandre.belloni@free-electrons.com> Acked-by:
Nicolas Ferre <nicolas.ferre@atmel.com> Cc: stable <stable@vger.kernel.org> Signed-off-by:
-
Ludovic Desroches authored
There is a macro named ATMEL_SERIAL_RINGSIZE which suggesting that it corresponds to the real size of the ring buffer. Let warn people that there is a factor of four since allocation size is sizeof(struct atmel_uart_char) * ATMEL_SERIAL_RINGSIZE. Signed-off-by:
Ludovic Desroches <ludovic.desroches@atmel.com> Signed-off-by:
-
Nicolas Ferre authored
Split TX and RX paths to not schedule RX tasklet on TX events and vice versa. Signed-off-by:
-
Nicolas Ferre authored
The IRQ status check and related actions was done in the tasklet without benefit. So, move it back to the IRQ context to simplify IRQ handling and having the possibility to split the tasklet in two separated ones for receive and transmit actions. Signed-off-by:
-
Muhammad Falak R Wani authored
Use memdup_user to duplicate a memory region from user-space to kernel-space, instead of open coding using kmalloc & copy_from_user. Signed-off-by:
Muhammad Falak R Wani <falakreyaz@gmail.com> Reviewed-by:
Samuel Thibault <samuel.thibault@ens-lyon.org> Signed-off-by:
-
Fabio Estevam authored
Doing the following UART bind/unbind sequence on a i.mx platform causes a kernel crash due to NULL pointer dereference: echo 21f4000.serial > /sys/bus/platform/drivers/imx-uart/bind echo 21f4000.serial > /sys/bus/platform/drivers/imx-uart/unbind Fix this problem by adding NULL checks prior to calling release/request_port ops. Reported-by:
Fugang Duan <fugang.duan@nxp.com> Signed-off-by:
Fabio Estevam <fabio.estevam@nxp.com> Tested-by:
-
Geert Uytterhoeven authored
Enabling support for the UART on the ARM MPS2 platform only make sense when compiling for the ARM MPS2 platform, unless compile-testing. Signed-off-by:
Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by:
Vladimir Murzin <vladimir.murzin@arm.com> Signed-off-by:
-
Ben Dooks authored
The msm_stop_dma() is not exported from the driver, so make it static to stop the following warning: drivers/tty/serial/msm_serial.c:84:6: warning: symbol 'msm_stop_dma' was not declared. Should it be static? Signed-off-by:
Ben Dooks <ben.dooks@codethink.co.uk> Acked-by:
Andy Gross <andy.gross@linaro.org> Reviewed-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Reviewed-by:
Stephen Boyd <sboyd@codeaurora.org> Signed-off-by:
-
Bjorn Andersson authored
For dm uarts in pio mode tx data is transferred to the fifo register 4 bytes at a time, but care is not taken when these 4 bytes spans the end of the xmit buffer so the loop might read up to 3 bytes past the buffer and then skip the actual data at the beginning of the buffer. Fix this by, analogous to the DMA case, make sure the chunk doesn't wrap the xmit buffer. Fixes: 3a878c43 ("tty: serial: msm: Add TX DMA support") Cc: Ivan Ivanov <iivanov.xz@gmail.com> Cc: stable@vger.kernel.org Reported-by:
Frank Rowand <frowand.list@gmail.com> Reported-by:
Nicolas Dechesne <nicolas.dechesne@linaro.org> Signed-off-by:
Frank Rowand <frank.rowand@am.sony.com> Reviewed-by:
-
Charanya Venkatraman authored
Avoid data corruption issues that result in CRC errors during file transfers over serial ports at higher baud rates. The current msm_serial driver masks the FIFO Rx interrupts in msm_start_rx_dma() since Rx FIFO interrupts are not required in DMA mode. However, msm_complete_rx_dma() re-enables the Rx FIFO interrupts which could cause RXSTALE event to be processed when a TXLEV interrupt occurs. The following is the sequence of events that could occur resulting in data corruption. msm_start_rx_dma -> msm_complete_rx_dma --> spin_unlock_irqrestore(&port->lock) --> msm_uart_irq()(For TXLEV interrupt) --> msm_handle_rx_dm() (Read from FIFO resulting in data corruption) The patch fixes the issue by not restoring the RXLEV and RXSTALE interrupts in msm_complete_rx_dma(). These interrupts are required only in FIFO mode. Signed-off-by:
Charanya Venkatraman <charanya@codeaurora.org> Acked-by:
-
Stephen Boyd authored
The hrtimer include isn't used and neither is serial. Drop those ones. The irq.h header really should be interrupt.h because this is an interrupt user and not an interrupt chip. Finally add wait.h for the wake_up*() usage in this driver and kernel.h for container_of(). Signed-off-by:
-
Stephen Boyd authored
The registers that msm_serial_set_mnd_regs() writes only exist on the non-uartdm hardware, so let's return early here if this function is called on uartdm hardware. This also prevents us from messing up the uartclk variable if the uartclk rate happens to be 19.2 or 4.8 MHz. Signed-off-by:
-
Stephen Boyd authored
This header file is only used by the driver, so let's merge the two together to reduce files and make it easier to see the whole driver without flipping through two files. This also makes it easier to use the structures defined in msm_serial.c in the functions that are defined in msm_serial.h by placing them in the proper locations. Signed-off-by:
-
Kefeng Wang authored
Some board like Hisilicon D02 uses Synopsys DesignWare ABP UART, declare an OF early console for it, so early console device can be enabled with comand line "earlycon"(without option) via the "stdout-path" property in device-tree. Signed-off-by:
Kefeng Wang <wangkefeng.wang@huawei.com> Tested-by:
Jon Mason <jon.mason@broadcom.com> Signed-off-by:
-
Geert Uytterhoeven authored
Add the missing timeout bit definition for (H)SCIF. Clear the timeout and overrun flag bits during UART reset, cfr. the initialization flowchart in the datasheet. Signed-off-by:
-
Geert Uytterhoeven authored
Setting the FIFO reset bits is not sufficient to reset the RX FIFO. After this the status register's RDF flag bit may still be set, causing the reception of one stale byte of data. To fix this, clear all status flag bits related to reception, error, and break handling, cfr. the initialization flowchart in the datasheet. Signed-off-by:
-
Geert Uytterhoeven authored
FIFO reset is done in sci_reset(), called from sci_set_termios(), while sci_start_tx() and sci_start_rx() are called before, from sci_startup(). However, starting transfers before the UART's FIFOs have been reset may cause reading of stale data. Remove the calls to sci_start_tx() and sci_start_rx() from sci_startup() to fix this. Transfers are still started when needed: - sci_start_rx() is called from sci_set_termios() after FIFO reset, if the CREAD flag is set, - sci_start_tx() is called from uart_change_speed() immediately thereafter, if transmission is enabled. Signed-off-by: -
Hiromitsu Yamasaki authored
Signed-off-by:
Hiromitsu Yamasaki <hiromitsu.yamasaki.ym@renesas.com> Signed-off-by:
Simon Horman <horms+renesas@verge.net.au> Acked-by:
-
Geert Uytterhoeven authored
Add support for indicating the availability of dedicated lines for RTS/CTS hardware flow control, using the standard "uart-has-rtscts" DT property. Signed-off-by:
-
Geert Uytterhoeven authored
The existing support for hardware-assisted RTS/CTS is rudimentary and doesn't work. Add support for hardware-assisted RTS/CTS hardware flow control for the (H)SCIF, SCIFA, and SCIFB variants. Signed-off-by:
-
Geert Uytterhoeven authored
Before, the driver relied on initialization by the boot loader, or by implicit reset state. Note that unlike on (H)SCIF, the RTS/CTS bits exist only if dedicated RTS/CTS pins are available, which depends on the SoC and UART instance. Signed-off-by:
-
Geert Uytterhoeven authored
Correct pin initialization on (H)SCIF: - RTS must be deasserted (it's active low), - SCK must be an input, as it may be used as the optional external clock input. Initial pin configuration must always be done: - Regardless of the presence of dedicated RTS and CTS pins: if the register exists, the RTS/CTS bits exist, too, - Regardless of hardware flow control being enabled or not: RTS must be deasserted. Signed-off-by: -
Geert Uytterhoeven authored
Improve documentation for the SCIFA/SCIFB Serial Port Control and Data Registers: - State clearly that the RTS and CTS lines are active-low, - Document the bits related to the serial port's SCK, RXD, and TXD pins. Signed-off-by:Peter Hurley <peter@hurleysoftware.com> Signed-off-by:
-
Geert Uytterhoeven authored
Improve documentation for the (H)SCIF Serial Port Register: - Make it clear the RTS and CTS lines are active-low, - Document the bits related to the serial port's clock pin. Signed-off-by:
-
Geert Uytterhoeven authored
Replace open-coded variants of sci_getreg() by function calls, and drop intermediate variables where appropriate. Signed-off-by:
-
Geert Uytterhoeven authored
Enhance the Renesas SCI UART driver to add support for GPIO-controlled modem lines (CTS, DSR, DCD, RNG, RTS, DTR), using the serial_mctrl_gpio helpers. GPIO-controlled modem lines can be used when dedicated modem lines are not available. Invalid configurations specifying both GPIO RTS/CTS and dedicated RTS/CTS are rejected. Signed-off-by:
-
Geert Uytterhoeven authored
Documentation/serial/driver clearly states: If the port does not support CTS, DCD or DSR, the driver should indicate that the signal is permanently active. Hence always set TIOCM_CTS, as we currently don't look at the CTS hardware line state at all. FWIW, this fixes the transmit path when hardware-assisted flow control is enabled, and userspace enables CRTSCTS. The receive path is still broken, as RTS is never asserted. Signed-off-by: -
Geert Uytterhoeven authored
Some Renesas SCIF UARTs have dedicated lines for RTS/CTS hardware flow control. Whether these lines exist depends on SoC and UART instance inside the SoC. Whether these lines can be used for hardware flow control depends on board wiring. Amend the DT bindings with an optional property to indicate that RTS/CTS hardware flow control lines exist, and can be used as such, according to the Generic Serial DT Bindings. Signed-off-by:
Rob Herring <robh@kernel.org> Cc: devicetree@vger.kernel.org Signed-off-by:
-
Geert Uytterhoeven authored
Amend the DT bindings for the Renesas SCI driver to allow describing optional GPIO-controlled modem lines, which can be used where dedicated modem lines are not available. The property naming is dictated by the Generic Serial DT Bindings. Signed-off-by:
-
Jiri Slaby authored
vc_deccolm is only set and never read, remove the member from vc_data. Signed-off-by:
Jiri Slaby <jslaby@suse.cz> Signed-off-by:
-
Jiri Slaby authored
We do not do hashtables for unicode fonts since 1995 (1.3.28). So it is time to remove the second parameter of con_clear_unimap and ignore the advice from userspace completely. Signed-off-by:
-
Jiri Slaby authored
Namely convert: * IS_FG -> con_is_fg * DO_UPDATE -> con_should_update * CON_IS_VISIBLE -> con_is_visible DO_UPDATE was a weird name for a yes/no answer, so the new name is con_should_update. Signed-off-by:
-
Jiri Slaby authored
Flush the switch cases to be aligned with the switch. Mostly everything can now fit to the 80-chars terminal. Signed-off-by:
-
Jiri Slaby authored
Do not opencode max3, use the macro. Separate commands. Until now, I have not noticed the comma. Make it one line, one command. And make the code obvious. Signed-off-by:
-
Jiri Slaby authored
The compiler noticed passing structure over stack. Even though rgb is a small structure, let us define one and pass that over all the functions wherever needed. Signed-off-by:
-
