- 09 Aug, 2010 38 commits
-
-
Jarod Wilson authored
Fix when CONFIG_MODULES is not enabled: drivers/staging/lirc/lirc_parallel.c:243: error: implicit declaration of function 'module_refcount' drivers/staging/lirc/lirc_it87.c:150: error: implicit declaration of function 'module_refcount' drivers/built-in.o: In function `it87_probe': lirc_it87.c:(.text+0x4079b0): undefined reference to `init_chrdev' lirc_it87.c:(.text+0x4079cc): undefined reference to `drop_chrdev' drivers/built-in.o: In function `lirc_it87_exit': lirc_it87.c:(.exit.text+0x38a5): undefined reference to `drop_chrdev' Its a quick hack and untested beyond building, since I don't have the hardware, but it should do the trick. Acked-by:
Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by:
Jarod Wilson <jarod@redhat.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@redhat.com>
-
Janne Grunau authored
Signed-off-by:
Janne Grunau <j@jannau.net> Signed-off-by:
-
Jarod Wilson authored
The ir_input_dev gets filled in by __ir_input_register, the one allocated in mceusb_init_input_dev was being overwritten by the correct one shortly after it was initialized (ultimately resulting in a memory leak). This bug was inherited from imon.c, and was pointed out to me by Maxim Levitsky. v2: fix incorrect dev arg to dev_dbg Signed-off-by:
-
Jarod Wilson authored
The ir_input_dev gets filled in by __ir_input_register, the one allocated in imon_init_idev was being overwritten by the correct one shortly after it was initialized (ultimately resulting in a memory leak). Additionally, there was an ill-advised memcpy into that extraneous ir_input_dev which gets fixed by this. Ill-advised memcpy pointed out by Dmitry Torokhov, bad usage of ir_input_dev pointed out by Maxim Levitsky. Signed-off-by:
-
Jarod Wilson authored
Per Dmitry Torokhov, following input_unregister_device with an input_free_device is forbidden, the former is sufficient alone. Acked-by:
Dmitry Torokhov <dtor@mail.ru> Signed-off-by:
-
Andy Walls authored
Signed-off-by:
Andy Walls <awalls@md.metrocast.net> Signed-off-by:
-
Andy Walls authored
The CX23885 and CX25840 modules were using their own simple IR pulse width measurement record type which required conversion when passing to the new IR core. This change makes that record type consistent with the new IR core and removes a data conversion. Signed-off-by:
-
Andy Walls authored
Instead of reporting an IR Rx timeout event as a ridiculously long space, report it as a space of the lenght of the timeout. This partially fixes operation with LIRC without breaking interoperation with the in kernel decoders. The gaps lengths reported to LIRC are still not real however. Signed-off-by:
-
Andy Walls authored
Compute and report the maximum IR pulse measurment width, even if we are set to perform carrier modulation or demodulation and the number is fixed by the carrier freq. Signed-off-by:
-
Andy Walls authored
This driver no longer depends on the old IR functions selected by VIDEO_IR. Signed-off-by:
-
Andy Walls authored
The CX23885 IR controller was reported to cause an interrupt storm on a TeVii S470 card, but was reported fine on an HVR-1250. Keep integrated IR disabled by default on CX2388[57] based cards to avoid a bad user experience in the general case. Signed-off-by:
-
Andy Walls authored
Interrupts from the AV Core are best handled by a workqueue handler since many I2C transactions are required to service the AV Core interrupt. The AV_CORE PCI interrupt is disabled by the IRQ handler and reenabled when the work handler is finished. Signed-off-by:
-
Andy Walls authored
This patch encapsulates access to the PCI_INT_MSK register and dev->pci_irqmask variable and protects them with a spinlock. This is needed because both the hard IRQ handler and a workhandler will need to manipulate the mask to disable the AV_CORE interrupt. Signed-off-by:
-
Andy Walls authored
Add initial IR Rx support using the intergrated IR controller in the A/V core of the CX23885 bridge chip. This initial support is flawed in that I2C transactions should not be performed in a hard irq context. That will be fixed in a follow on patch. The TeVii S470 support is reported to generate perptual interrupts that renders a user' system nearly unusable. The TeVii S470 IR will be disabled by default in a follow on patch. Signed-off-by:
-
Andy Walls authored
Signed-off-by:
-
Andy Walls authored
This patch is primarily a port of the integrated IR controller code in cx23885/cx23888-ir.c. Right now, only the CX2388[57] AV core will really try to set up IR. This IR support, by design, still requires the bridge driver to do final IO pin mux configuration and setup of the IR controller parameters. For the CX2388[578] chips, enabling the AV Core for IR also starts sending Audio and Video interrupts to the bridge driver. For CX2388[578] chips audio and video interrupts are ignored and acknowledged when they happen. IR interrupt handling and status logging is exluded for the CX23888 which does not have an IR controller on the AV core. Note that experimentation reveals that the IR irq enables on the CX23885 have an inverted logic sense. The CX23887 likely suffers from the same quirk. For these chips, those irq enable bits are handled as interrupt disables. Signed-off-by:
-
Andy Walls authored
The CX2584x and related cores are multifunction subdevices with a number of internal blocks that act as interrupt sources. Move the v4L2_subdev interrupt_service_routine callback from v4l_subdev_ir_ops to v4l2_subdev_core_ops, as the video and audio blocks of a CX2584x and related cores can generate interrupts along with the IR block. This change also makes sense for other subdev's that generate interrupts and do not have an IR block. Signed-off-by:
-
Andy Walls authored
Explicitly configure the IR Tx and IR Rx pins to be driven by the IR Rx and Tx pads from the AV core for CX23888 IR. For the HVR-1850 and HVR-1290 configure the IR Tx level inversion, so the Tx LED is off when idle. Signed-off-by:
-
Andy Walls authored
There is a distinction on IR Tx for the CX2388[578] chips of carrier sense inversion (space is a carrier burst and mark is no burst) and I/O pin level inversion (0 is high output level, 1 is low output level). Allow the caller to set these parameters distinctly as v4l2_subdevice IR parameters. This permits the IR device to be configured and enabled without the IR Tx LED being on during idle/space time due to an external hardware level inversion Signed-off-by:
-
Andy Walls authored
Add s_io_pin_config core subdev op for the CX2388[578] AV cores. This is complete for IR_RX, IR_TX, GPIOs 16,19-23, and IRQ_N. It likely needs work for the I2S signal direction. Signed-off-by:
-
Andy Walls authored
Add a method to v4l2_sudev_core_ops to allow bridge drivers to manage what signal pads/functions are routed out to multiplexed IO pins on a pin by pin basis. The interface also allows specifying initial output settings for pins and disabling an IO pin altogether. Signed-off-by:
-
Andy Walls authored
Add a simple log_status function for raw analog video capture device nodes, to provide insight into the state of the CX2388[578] A/V decoder core. Signed-off-by:
-
Andy Walls authored
The offset in the eeprom data for the 79501 version of the HVR-1250 is at 0xc0 vs. the standard 0x80. Signed-off-by:
-
Andy Walls authored
There was a small window between writing the cx25840 register address over the i2c bus and reading the register contents back from the cx25840 device that the i2c adapter lock was released. This change ensures the adapter lock is not released until the register read is done. Signed-off-by:
-
Jean Delvare authored
Function i2c_wait_done() never returns negative values, so there is no point in checking for them. Signed-off-by:
Jean Delvare <khali@linux-fr.org> Signed-off-by:
-
Jean Delvare authored
Don't just check for nacks on zero-length transactions. Check on other transactions too. Signed-off-by:
-
Jean Delvare authored
Documentation/i2c/fault-codes says that i2c adapter drivers should return -ENXIO when no slave acks an address byte. Signed-off-by:
-
Jean-François Moine authored
On laptops, the webcam sometimes disconnects when writing the first main register (b300). This patch prevents the creation of a new /dev/video<n> on streaming start. Signed-off-by:
Jean-François Moine <moinejf@free.fr> Signed-off-by:
-
Jean-François Moine authored
Signed-off-by:
-
Jean-François Moine authored
The added controls are exposure, gain, autogain and backlight compensation. Signed-off-by:
-
Jean-François Moine authored
Signed-off-by:
-
Jean-François Moine authored
Signed-off-by:
-
Jean-François Moine authored
The JPEG format did not work fine. The Bayer format offers correct VGA (640x480) resolution, but bad QQVGA (160x120). This last resolution is removed. Signed-off-by:
-
Jean-François Moine authored
Signed-off-by:
-
Jean-François Moine authored
The bit 7 of the sensor gain register multiplies the 7bits value by two. It is now always set with half the user gain value. Signed-off-by:
-
Jean-François Moine authored
The start_1 exchange must be done by (reg, val) * 5 and not 8. Signed-off-by:
-
Jean-François Moine authored
Signed-off-by:
-
Jean-François Moine authored
The sensor enum definition was not at the right place. Signed-off-by:
-
- 08 Aug, 2010 2 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tileLinus Torvalds authored
* git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile: arch/tile: check kmalloc() result arch/tile: catch up on various minor cleanups. arch/tile: avoid erroneous error return for PTRACE_POKEUSR. tile: set ARCH_KMALLOC_MINALIGN tile: remove homegrown L1_CACHE_ALIGN macro arch/tile: Miscellaneous cleanup changes. arch/tile: Split the icache flush code off to a generic <arch> header. arch/tile: Fix bug in support for atomic64_xx() ops. arch/tile: Shrink the tile-opcode files considerably. arch/tile: Add driver to enable access to the user dynamic network. arch/tile: Enable more sophisticated IRQ model for 32-bit chips. Move list types from <linux/list.h> to <linux/types.h>. Add wait4() back to the set of <asm-generic/unistd.h> syscalls. Revert adding some arch-specific signal syscalls to <linux/syscalls.h>. arch/tile: Do not use GFP_KERNEL for dma_alloc_coherent(). Feedback from fujita.tomonori@lab.ntt.co.jp. arch/tile: core support for Tilera 32-bit chips. Fix up the "generic" unistd.h ABI to be more useful.
-
git://www.jni.nu/crisLinus Torvalds authored
* 'for-linus' of git://www.jni.nu/cris: (51 commits) CRIS: Fix alignment problem for older ld CRIS: Always dump registers for segfaulting process. CRIS: Add config for pausing a seg-faulting process CRIS: Don't take faults while in_atomic CRIS: Fixup lookup for delay slot faults CRIS: Discard exit.text and .data at runtime CRIS: Add cache aligned and read mostly data sections CRIS: Return something from profile write CRIS: Add ARTPEC-3 and timestamps for sync-serial CRIS: Better ARTPEC-3 support for gpio CRIS: Add include guard CRIS: Better handling of pinmux settings CRIS: New DMA defines for ARTPEC-3 CRIS: __do_strncpy_from_user: Don't read the byte beyond the nil CRIS: Pagetable for ARTPEC-3 CRIS: Machine dependent memmap.h CRIS: Check if pointer is set before using it CRIS: Machine dependent dma.h CRIS: Define __read_mostly for CRISv32 CRIS: Discard .note.gnu.build-id section ...
-
