| Age | Commit message (Collapse) | Author |
|---|
|
Changes according to TVE-VGA patch.
Signed-off-by: Jason Chen <b02280@freescale.com>
(cherry picked from commit 80379f44148be8cb5a39f75e38586100f667d5db)
|
|
Add VGA support.
To enable VGA on second display, you can do the following:
> echo U:1280x1024p-60 > /sys/class/graphics/fb1/mode
> echo 0 > /sys/class/graphics/fb1/blank
To enable VGA on primary display, you can set bootcmd:
video=mxcdi1fb:BGR24,VGA-XGA di1_primary
or
video=mxcdi1fb:BGR24,SXGA di1_primary
Signed-off-by: Jason Chen <b02280@freescale.com>
(cherry picked from commit e6b393f8b9f32d0a982945ad6c9a54ec5f07923b)
|
|
Add VGA support for ipu basic driver and fb driver.
Signed-off-by: Jason Chen <b02280@freescale.com>
(cherry picked from commit 214d8593acc37e2d5ed32e75a6518a8e306a1246)
|
|
Remove the 2G limitation from NAND driver since currently kernel
can support more than 2GB NAND flash now. This commit will make
NAND driver rescan the whole NAND to create one BBT and store to
the last 4 blocks of the larger than 2GB NAND flash. This commit
will have no effect on the NAND which size is not larger than 2GB.
Signed-off-by: Jason Liu <r64343@freescale.com>
(cherry picked from commit 5790e8a3e8a3fabc4fe7ab5f4c3524f1791e073f)
|
|
As IPUv3 CSI supports interlaced scan order for CSI->SMFCx->MEM
channels, change TVin PAL/NTSC video formats to be standard.
Signed-off-by: Liu Ying <b17645@freescale.com>
(cherry picked from commit aef894228ba2c94cff1029a4cbd6b8e9a0a8d89b)
|
|
This patch supports interlaced scan order for CSI->SMFCx->MEM
channels.
Signed-off-by: Liu Ying <b17645@freescale.com>
(cherry picked from commit c3b90654b76654911a2e132dd387ef9a9422e114)
|
|
1) Add re-check to avoid event lost in edge irq mode
2) fix mutex problem caused by CHECK_ERROR(xxx) macro
Signed-off-by: Zhou Jingyu <Jingyu.Zhou@freescale.com>
(cherry picked from commit 4837fb8a06c8ffd959d20785cfaf4104689b4906)
|
|
This patch supports bottom field first video.
Signed-off-by: Liu Ying <b17645@freescale.com>
(cherry picked from commit b1123ead7b6da887c0139b455dbaf00780e7e0bb)
|
|
This patch supports bottom field first setting.
Signed-off-by: Liu Ying <b17645@freescale.com>
(cherry picked from commit c36dbf686a6e0dd7e7e746ad986105255de0b7c0)
|
|
If VPU doesn't use IRAM, there will be a kernel dump when
we enter VPU driver probe function.
This issue is caused by allocating IRAM with zero buffer
size.
Signed-off-by: Liu Ying <b17645@freescale.com>
(cherry picked from commit 020fa3509623f49aa759fb47864b70eb017403f8)
|
|
MX51: Enable battery charger on i.MX51 babbage.
The issue that board would be damaged if a battery was not connected
was fixed on MC13892 version 2.0
Signed-off-by: Nancy Chen <Nancy.Chen@freescale.com>
|
|
Not all the NAND devices have all the information in additional
id bytes.
So add a hook in the nand_chip{} is a good method to calculate the
right value of oobsize, erasesize and so on.
Without the hook,you will get the wrong value, and you have to hack
in the ->scan_bbt() to change the wrong value which make the code
mess.
Signed-off-by: Huang Shijie <shijie8@gmail.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
|
|
The ONFI nand needs the GF14.
So change the block mark's position for ONFI NAND.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
HW flow control should be enable/disable by userspace.
It can be set by tty option: options.c_cflag of CRTSCTS
bit.
This bug is found by testing UART BT AR3001 chip,
this chip will wake up the chip by set RTS to high,
it needs frist disable HW flow control, set RTS,
then disable HW flow control.
Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>
|
|
Fix dhcp failure at i.MX28 EVK board caused by 1588 patch
Signed-off-by: Xie Xiaobo <R63061@freescale.com>
Signed-off-by: Frank Li <Frank.Li@freescale.com>
|
|
when plug A-cable, but no usb device connect to the cable's receptacle,
then unplug the cable, and then plug the cable again, the wakeup
function can not work.
Signed-off-by: Hu Hui <b29976@freescale.com>
|
|
- Hang was caused by accessing EPDC registers without enabling clocks to the
EPDC module. Fix is to defer actual write to EPDC registers until subsequent
screen updates are requested.
Signed-off-by: Danny Nold <dannynold@freescale.com>
|
|
- User can specify a flag (EPDC_FLAG_FORCE_MONOCHROME) to have 8-bit grayscale
converted to monochrome (black or white) via processing in the PxP
(using the LUT).
- Added logic to check against the full videomode when identifying the
desired FB mode. This allows support for multiple modes with the same
dimensions.
Signed-off-by: Danny Nold <dannynold@freescale.com>
|
|
Driver Part
move the usb clk_enable from irq context to a kernel thread context, so
that the voltage can be changed in clk_enable function.
Signed-off-by: Hu Hui <b29976@freescale.com>
|
|
If tell the real correcting infomation to the upper layer of
MTD, the torture thread of UBIFS will do the torture test in
a very often frequency. This will eat up all the reservation blocks
of the UBIFS.
So tell the real correcting infomation only when the failure occured,
or the corrected times nearly reached the ECC threshold.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
The nfc->begin() or nfc->end() will do the relative clock
opertions. Enable clock or disable clock here is redundant.
So just remove it.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
fix compile warnings for the GPMI.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
In imx50, the low power feature is introduced into the system.
We have to enable the ddr_clk explicitly before doing any DMA operations.
So enable the ddr_clk in ->begin(), and disable it in ->end();
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
Add the ONFI nand init code.
In DDR mode of ONFI NAND, the data READ/WRITE will become 16-bit,
although the ALE/CLE are still use the 8-bit.
So changing the DMA command chains of ECC READ/WRITE page's operations
to support the ONFI nand.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
ONFI nand needs the 24-bit ECC and 1k data chunk.
DMA page size changes to the real used size.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
ONFI nand works in asynchronous mode in normal case,
You have to do extra initialization to let it work in
DDR mode(synchronous mode).
so add extra initialization hook for ONFI nand.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
ONFI nand needs 24-bit ECC per 1080 bytes of data.
So add a 24-bit ECC search table for the ONFI nand.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
Add the new field to distinguish the ONFI nand from the others.
Alos add the function is_onfi_nand() to check the ONFI nand.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
add the necessary id infomation for the MICRON ONFI nand.
Signed-off-by: Huang Shijie <b32955@freescale.com>
|
|
The Linux I2C framework expects the slave address as the 7LSB bits,
therefore the address must be shifted 1 to left to become the real
byte sending on the I2C bus.
Signed-off-by: Aisheng.Dong <b29396@freescale.com>
|
|
When v4l2 using ic bypass mode, and format is non-interleaving, if
the resolution is large, like 720P on mx51, it has chance to make
screen black flash.
Need enlarge burst size of IDMA display for above case.
Signed-off-by: Jason Chen <b02280@freescale.com>
Signed-off-by: Ferderber Ran <r53561@freescale.com>
|
|
Some variable change at 2.6.35. So ubiblock need update to use
new member variable.
Signed-off-by: Frank Li <Frank.Li@freescale.com>
|
|
register_blkdev return 1..255 when major = 0.
if (ret ) {
printk(KERN_WARNING "Unable to register %s block device on major %d: %d\n",
tr->name, tr->major, ret);
mutex_unlock(&mtd_table_mutex);
return ret;
}
Above code will return fail when register_blkdev return allocated major number.
Signed-off-by: Patrick Turley <patrick.turley@freescale.com>
Signed-off-by: Frank Li <Frank.Li@freescale.com>
|
|
Supply the 1588 support for MX28 FEC.
Signed-off-by: Xie Xiaobo <X.Xie@freescale.com>
|
|
This patch checks zero width/height parameters provided by user
in ioctrl VIDIOC_S_FMT. It can fix the issue of division by zero
in kernel in this case.
Signed-off-by: Liu Ying <b17645@freescale.com>
|
|
PMIC: Charger removal may not always be detected by MC13892.
Signed-off-by: Nancy Chen <Nancy.Chen@freescale.com>
|
|
Workaround for imx5x M4IF burst len errata when armv7 neon enabled, use
pgprot_writecombine instead of pgprot_noncached, to avoid burst len > 8.
Signed-off-by: Sammy He <r62914@freescale.com>
|
|
Backlight value is reset at event MXS_LCDIF_PANEL_INIT after resume.
Check suspend status to avoid reset value.
Signed-off-by: Frank Li <Frank.Li@freescale.com>
|
|
Though users can specify 'ldb=xxx' to enable 1080P LVDS
panel when using DI1, it is not easy for them to set it
correctly. This patch will add the default LDB setting
for this case and correct the comment of ldb_setup()
function.
Signed-off-by: Liu Ying <b17645@freescale.com>
|
|
This patch makes sure that slave device is present when opening
related video device. It will avoid a kernel dump if slave device
is not present before calling custom open function.
Signed-off-by: Liu Ying <b17645@freescale.com>
|
|
Fix build error at 2.6.25 kernel.
wrong write\read register use structure without volatile.
Interrupt handle will dead loop because complier optimize.
Kernel perfer use readl and writel to access register.
Change all register access to readl\writel.
Signed-off-by: Frank Li <Frank.Li@freescale.com>
|
|
The L2 Switch can get mac address from platform data.
Signed-off-by: Xie Xiaobo <X.Xie@freescale.com>
|
|
The L2 Switch have 4 ports,
and the port0 can be used as ethernet port
when Switch is configured to operate
as a 3-Port Switch (Switch Mode).
Signed-off-by: Niu Xule <b23300@freescale.com>
|
|
L2 switch driver command line controlling support
Signed-off-by: Niu Xule <b23300@freescale.com>
|
|
L2 Switch support for mx28
Signed-off-by: Niu Xule <b23300@freescale.com>
|
|
Do not use spin_lock when sets power on/off since pmic driver may sleep
on this operation.
Signed-off-by: Aisheng.Dong <b29396@freescale.com>
|
|
Previous patch is not stable enough, reverted.
Signed-off-by: Aisheng.Dong <b29396@freescale.com>
|
|
if enable the auto suspend for the usb device attached to the usb otg
host port, after the usb device and it's roothub,hcd auto suspend,
disconnecting the usb device can not be realized by the roothub, because
the roothub is in the suspend state.
Signed-off-by: Hu Hui <b29976@freescale.com>
|
|
Kobs-ng use MTD_MODE_RAW write boot stream to NAND flash.
mtdchar have not initilized all field of struct mtd_oob_ops
in mtd_write when mode is MTD_MODE_RAW.
function nand_do_write_ops
if (ops->ooboffs && (ops->ooboffs + ops->ooblen > oobmaxlen))
return -EINVAL;
ooboffs is not initilized at mtd_write.
So random data cause nand_do_write_ops return fail.
Signed-off-by: Frank Li <Frank.Li@freescale.com>
|
|
when doing write barrier, the data of the bio is NULL, it's OK for ramdisk
doing nothing, So the result of the bio must be success.
Signed-off-by: Hu Hui <b29976@freescale.com>
|
