1. dw_i2c for hw not enable IC_EMPTYFIFO_HOLD_MASTER_EN,RX FIFO may
over if SW not read RX FIFO day timely.So enabel RX OVER INT,when
receive RX OVER ,i2c_dw_xfer return -EAGAIN,to trigger retry.
2. for large i2c read, set DW_IC_RX_TL to rx_fifo_depth / 2, to make more
efficient irq process.
3.skip isr once rx over by disable INT
4. merge fix from upstream
commit 301c8f5c32c8fb79c67539bc23972dc3ef48024c
Author: Jarkko Nikula <jarkko.nikula@linux.intel.com>
Date: Tue Sep 27 16:56:44 2022 +0300
i2c: designware: Fix handling of real but unexpected device interrupts
Commit c7b79a752871 ("mfd: intel-lpss: Add Intel Alder Lake PCH-S PCI
IDs") caused a regression on certain Gigabyte motherboards for Intel
Alder Lake-S where system crashes to NULL pointer dereference in
i2c_dw_xfer_msg() when system resumes from S3 sleep state ("deep").
I was able to debug the issue on Gigabyte Z690 AORUS ELITE and made
following notes:
- Issue happens when resuming from S3 but not when resuming from
"s2idle"
- PCI device 00:15.0 == i2c_designware.0 is already in D0 state when
system enters into pci_pm_resume_noirq() while all other i2c_designware
PCI devices are in D3. Devices were runtime suspended and in D3 prior
entering into suspend
- Interrupt comes after pci_pm_resume_noirq() when device interrupts are
re-enabled
- According to register dump the interrupt really comes from the
i2c_designware.0. Controller is enabled, I2C target address register
points to a one detectable I2C device address 0x60 and the
DW_IC_RAW_INTR_STAT register START_DET, STOP_DET, ACTIVITY and
TX_EMPTY bits are set indicating completed I2C transaction.
My guess is that the firmware uses this controller to communicate with
an on-board I2C device during resume but does not disable the controller
before giving control to an operating system.
I was told the UEFI update fixes this but never the less it revealed the
driver is not ready to handle TX_EMPTY (or RX_FULL) interrupt when device
is supposed to be idle and state variables are not set (especially the
dev->msgs pointer which may point to NULL or stale old data).
Introduce a new software status flag STATUS_ACTIVE indicating when the
controller is active in driver point of view. Now treat all interrupts
that occur when is not set as unexpected and mask all interrupts from
the controller.
Change-Id: If81c92a03b9be8699742184b46c03df34f922a28
1.broken-cd: The default non-removable attribute fails to recognize the AIC8800, thus it is set to rely on a software card detection signal instead of hardware, always treating the card slot as if a card is present.
2.keep-power-in-suspend: The SDIO of TH1520 does not support keeping power when entering STR, and the driver of AIC8800 does not support SDIO with this function. Therefore, this node is added here to meet the driver requirements of AIC8800.
3.support_power_ctrl:The newly added device node is used to control the WL-REG-ON signal (which controls the power of AIC8800) after waking up from STR.
Change-Id: Ia33309da9f3e6bdf291f408d26909f11afd24842
This commit fixup unsafe release of allocater and dmabuf.
have merged
vidmem:kernel: update dmabuf release not free by allocate release.
Prev commit add release of dmabuf may caused mem node list
removed racing.Here have moved the exported node to global
list.
Signed-off-by: xianbing Zhu <xianbing.zhu@linux.alibaba.com>
Change-Id: I8c2a83b60f7d6ff26771550cc691db6b27b73609
Signed-off-by: Han Gao <gaohan@iscas.ac.cn>
This commit implement dma_buf_ops release to free buffer
in function _dmabuf_release(struct dma_buf *dmabuf).
Default use pr_debug for dynamic debug in macro DEBUG_PRINT.
Signed-off-by: xianbing Zhu <xianbing.zhu@linux.alibaba.com>
Change-Id: Id0a5841ccf8459c469034866dd2878e8698a57a4
Signed-off-by: Han Gao <gaohan@iscas.ac.cn>
This issue is reported in TLV-299.
This commit fixed allocate_cmdbuf may infinite loop
and leak of release the cmdbuf only reserved but not
link_and_run.
If cmdbuf not run done when release,do not firstly abort,
instead wait done event reduce abort failed.
Add cmdbuf id and line info for error log.
Fixed busy count record idle twice in read module all
regs issue.
Signed-off-by: xianbing Zhu <xianbing.zhu@linux.alibaba.com>
Change-Id: Ia006f8123e38022400959bed838ef76f7aef14a0
Signed-off-by: Han Gao <gaohan@iscas.ac.cn>
This issue is reported in TLV-299.
This commit fixed allocate_cmdbuf may infinite loop
and leak of release the cmdbuf only reserved but not
link_and_run.
If cmdbuf not run done when release,do not firstly abort,
instead wait done event reduce abort failed.
Add cmdbuf id and line info for error log.
Fixed busy count record idle twice in read module all
regs issue.
Signed-off-by: xianbing Zhu <xianbing.zhu@linux.alibaba.com>
Change-Id: Ieb0b751d18a1ed65fe198e3d75a6953dabf15ef8
Signed-off-by: Han Gao <gaohan@iscas.ac.cn>
Fixed the issue where HDMI maximum resolution is 1024x768.
With this commit added, the maximum HDMI resolution is 4K.
Signed-off-by: Haaland Chen <haaland@milkv.io>
check-config.sh script of k3s kubernetes light distribution flags the following configs as missing:
CONFIG_NETFILTER_XT_MATCH_COMMENT
CONFIG_NETFILTER_XT_MATCH_MULTIPORT
CONFIG_IP_SET
CONFIG_INET_XFRM_MODE_TRANSPORT
It fixes the issue that u-boot failed to load dtbo like this:
Retrieving file: /dtbs/linux-image-5.10.113-lpi4a/thead/overlays/meles-wifibt-external-antenna.dtbo
715 bytes read in 4 ms (173.8 KiB/s)
failed on fdt_overlay_apply(): FDT_ERR_NOTFOUND
base fdt does did not have a /__symbols__ node
make sure you've compiled with -@
Failed to apply overlay /dtbs/linux-image-5.10.113-lpi4a/thead/overlays/meles-wifibt-external-antenna.dtbo, skipping
ERROR: Did not find a cmdline Flattened Device Tree
Loading Ramdisk to 1fa22000, end 1ffff820 ... OK
Device tree not found or missing FDT support
### ERROR ### Please RESET the board ###
Signed-off-by: Haaland Chen <haaland@milkv.io>
