[V2][linux-yocto][linux-yocto v5.2/standard/base][linux-yocto-dev standard/base][PATCH 0/1] yaffs: add strict check when call yaffs_internal_read_super

From: Quanyang Wang Hi Bruce, V1 ---> V2: 1. split into 2 patches to make the fix more reasonable. 2. add crash info to the patch. Would you please help merge this patch to the branches for: linux-yocto v5.2/standard/base linux-yocto-dev standard/base Quanyang Wang (2): yaffs: repair

[linux-yocto][PATCH 2/9] arm: zynq: delete AFLAGS_suspend.o to fix compile warning

From: Quanyang Wang When using arm-linux-gnueabihf-gcc v9.2.0 to compile kernel, there will be warning info as below: cc1: warning: switch '-mcpu=cortex-a9' conflicts with '-march=armv7-a' switch This is because that if "-march" and "-mcpu" options are passed to gcc together, gcc will check if

[linux-yocto][PATCH 6/9] drivers: soc: xilinx: fix compile warning by defining payload array length to 32

From: Quanyang Wang In zynqmp-ipi-mailbox.c and zynqmp_r5_remoteproc.c, the variables of zynqmp_ipi_message structures are all using 32 as the maximum size of data length. So move IPI_BUF_LEN_MAX to zynqmp-ipi-message.h to share with zynqmp_power.c. Signed-off-by: Quanyang Wang Acked-by:

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 0/9] Patches from linux-yocto v5.2 branch

From: Quanyang Wang Hi Bruce & Michal, There are 9 patches in this pull request. The patch "0001-ARM-dts-zynq-enablement-of-coresight-topology.patch" is picked from the upstream commit from Michal and delete ITM compatible string from it. The patch

[linux-yocto][PATCH 4/9] clk: zynqmp: fix memory leak in zynqmp_register_clocks

From: Quanyang Wang This is detected by kmemleak running on zcu102 board: unreferenced object 0xffc877e48180 (size 128): comm "swapper/0", pid 1, jiffies 4294892909 (age 315.436s) hex dump (first 32 bytes): 64 70 5f 76 69 64 65 6f 5f 72 65 66 5f 64 69 76 dp_video_ref_div 31 00 00 00 00 00

[linux-yocto][PATCH 8/9] sound: soc: xilinx: give a name to stream_name in xilinx_dp_dai_links

From: Quanyang Wang If .stream_name is NULL, when calling "aplay -l", it will show: card 0: monitor [DisplayPort monitor], device 0: (null) xilinx-dp-snd-codec-dai-0 [(null) xilinx-dp-snd-codec-dai-0] Adding .stream_name can eliminate the "(null)" in the output. Signed-off-by: Quanyang Wang

[linux-yocto][PATCH 7/9] i2c: cadence: keep bus_hold_flag unless I2C_M_NOSTART is set

From: Quanyang Wang When using i2c_smbus_read_byte_data to read one byte from a slave device, because of the commit d358def70688 ("i2c: cadence: Fix the hold bit setting"), the transaction becomes: S Addr Wr [A] Comm [A] P S Addr Rd [A] [Data] NA P ^

[linux-yocto][PATCH 1/9] ARM: dts: zynq: enablement of coresight topology

From: Zumeng Chen This patch is to build the coresight topology structure of zynq-7000 series according to the docs of coresight and userguide of zynq-7000. Signed-off-by: Zumeng Chen Link: https://patchwork.kernel.org/patch/11181257/ [Quanyang: delete compatible string for itm node]

[linux-yocto][PATCH 1/1] yaffs: add strict check when call yaffs_internal_read_super

From: Quanyang Wang When kernel booting, mount_block_root will be called to judge the filesystem type of root device. Then .mount function in file_system_type structure will do the check operation. But yaffs filesystem has a relaxed examination because as a filesystem for NAND Flash, it doesn't

[linux-yocto][PATCH 5/9] dma: xilinx: dpdma: alloc xilinx_dpdma_tx_desc using GFP_ATOMIC

From: Quanyang Wang The function xilinx_dpdma_chan_alloc_tx_desc can be called from atomic context, hence use GFP_ATOMIC instead of GFP_KERNEL, or else it will cause the calltrace as below: BUG: sleeping function called from invalid context at mm/slab.h:418 in_atomic(): 1, irqs_disabled(): 128,

[linux-yocto][PATCH 3/9] ARM: zynq: use physical cpuid in zynq_slcr_cpu_stop/start

From: Quanyang Wang When kernel booting, it will create a cpuid map between the logical cpus and physical cpus. In a normal boot, the cpuid map is as below: Physical Logical 0==> 0 1==> 1 But in kdump, there is a condition that the crash happens at the

[linux-yocto][PATCH 9/9] xilinx-zynq: Revert "zynq: SMP crashkernel support"

From: Quanyang Wang This reverts commit 5d1a5900882ceb86cddf36c258bd5bda498c8da8. Since the commit ("ARM: zynq: use physical cpuid in zynq_slcr_cpu_stop/start") has fixed the hang issue in kdump, this patch can be reverted. Signed-off-by: Quanyang Wang Acked-by: Michal Simek Signed-off-by:

Re: [linux-yocto][PATCH 1/1] yaffs: add strict check when call yaffs_internal_read_super

On 12/3/19 4:38 PM, Kevin Hao wrote: On Tue, Dec 03, 2019 at 04:27:15PM +0800, quanyang.wang wrote: From: Quanyang Wang When kernel booting, mount_block_root will be called to judge the filesystem type of root device. Then .mount function in file_system_type structure will do the check

[linux-yocto][PATCH 2/3] xilinx-zynqmp: add xlnx_drm_framebuffer_init to replace drm_gem_fbdev_fb_create

From: Quanyang Wang Since the function drm_gem_fbdev_fb_create has been removed in commit 73415b4ea5b5 ("drm/fb: remove unused function: drm_gem_fbdev_fb_create()"), add a new function xlnx_drm_framebuffer_init to create a drm_framebuffer structure. Signed-off-by: Quanyang Wang ---

[linux-yocto] [linux-yocto-dev standard/xlnx-soc]][PATCH 0/3] New patches for v5.4

From: Quanyang Wang Hi Bruce & Michal, There are 3 patches in this pull-request. 1) The patch "0001-spi-zynq-qspi-use-wait_for_completion_timeout-instea.patch" fixes a bug which occurs in zc706 and zc702 boards using mainline and linux-yocto kernel. Using the command as below can reproduce

[linux-yocto][PATCH 3/3] spi: spi-mem: zynq-qspi: add is-dual support for zc706 board

From: Quanyang Wang There are 2 SPI flash memories in zc706 board and it use the configuration "Dual SS, 8-bit Parallel" which means separate Slave-Select lines and separate 4 data lines. So add the is-dual support for zc706 board according to the spi driver in SDK. Refer to bd2c1810ae87 ("spi:

[linux-yocto][PATCH 1/3] spi: zynq-qspi: use wait_for_completion_timeout instead of wait_for_completion_interruptible_timeout

From: Quanyang Wang When umount jffs2 filesystem on mtdblock device, the process "jffs2_gcd_mtdX" which calls zynq_qspi_exec_mem_op to read/write flash will receive signal "SIGKILL" from thread "jffs2_stop_garbage_collect_thread". This means that at this moment the function

[linux-yocto][linux-yocto v5.2/standard/base][linux-yocto-dev standard/base][PATCH 0/1] yaffs: add strict check when call yaffs_internal_read_super

From: Quanyang Wang Hi Bruce, Would you please help merge this patch to the branches for: linux-yocto v5.2/standard/base linux-yocto-dev standard/base This patch fixes a bug: when using MMC card with btrfs filesystem to be root device, kernel will crash and the log as below: mmc0: new high

[linux-yocto][PATCH 1/2] yaffs: repair yaffs_get_mtd_device

From: Quanyang Wang The function yaffs_get_mtd_device use wrong function to retrieve mtd_info structure (using yaffs_get_mtd_device itself will cause dead loop). Use get_mtd_device to do this. Signed-off-by: Quanyang Wang --- fs/yaffs2/yaffs_mtdif.c | 12 ++-- 1 file changed, 10

[linux-yocto][PATCH 2/2] yaffs: add strict check when call yaffs_internal_read_super

From: Quanyang Wang When kernel booting, mount_block_root will be called to judge the filesystem type of root device. Then .mount function in file_system_type structure will do the check operation. But yaffs filesystem has a relaxed examination because as a filesystem for NAND Flash, it doesn't

Re: [linux-yocto][PATCH 2/3] xilinx-zynqmp: add xlnx_drm_framebuffer_init to replace drm_gem_fbdev_fb_create

Hi Hyun, On 12/13/19 2:10 AM, Hyun Kwon wrote: Hi Quanyang, Thanks for the patch. -Original Message- From: Michal Simek Sent: Thursday, December 12, 2019 5:13 AM To: quanyang.w...@windriver.com; Bruce ; Michal Simek ; Hyun Kwon Cc: linux-yo...@yoctoproject.org Subject: Re:

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc]][PATCH 0/3] New patches for v5.4

Hi Michal, Any comment? Thanks, Quanyang On 12/3/19 6:11 PM, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce & Michal, There are 3 patches in this pull-request. 1) The patch "0001-spi-zynq-qspi-use-wait_for_completion_timeout-instea.patch" fixes a bug which occurs in

Re: [linux-yocto][PATCH 3/3] spi: spi-mem: zynq-qspi: add is-dual support for zc706 board

Hi Naga, On 12/13/19 6:29 PM, Naga Sureshkumar Relli wrote: Hi Quanyang, The driver zynq-qspi.c which was written under spi-mem framework supports only single CS. The changes you are done to support dual-parallel, inside the driver are fine from controller side but there Is no support yet in

[linux-yocto][PATCH 1/2] xilinx-zynq: add the support for xlinx-zynq bsp

From: Quanyang Wang This add cfg files to support xilinx zc702 and zc706 boards. Signed-off-by: Quanyang Wang Signed-off-by: Bruce Ashfield --- bsp/xilinx-zynq/xilinx-zynq-standard.scc | 7 + bsp/xilinx-zynq/xilinx-zynq.cfg | 199 +++

[linux-yocto][kernel-cache master] xilinx-zynq: add xilinx-zynq kernel-cache for master branch

From: Quanyang Wang Hi Bruce, These patches are picked up from kernel-cache v5.2 branch, and no modification. Would you please help merge the patches to yocto-kernel-cache master branch? Quanyang Wang (2): xilinx-zynq: add the support for xlinx-zynq bsp xilinx-zynq: enable coresight and

[linux-yocto] [V2] [linux-yocto-dev standard/xlnx-soc] New patches for v5.4

From: Quanyang Wang Hi Bruce & Michal, Would you please help review and merge these patches to linux-yocto-dev standard/xlnx-soc branch? There are 9 patches in this V2 pull request. I drop all V1 patches and will use the patches as below: The patches:

[linux-yocto][PATCH 6/9] spi: zynq-qspi: Add driver for zynq qspi

From: Michal Simek The driver supports single, dual parallel and dual stacked configurations. Signed-off-by: Naga Sureshkumar Relli Signed-off-by: Michal Simek [Quanyang: original patch taken from https://github.com/Xilinx/linux-xlnx.git xlnx_rebase_v4.19] Signed-off-by: Quanyang Wang ---

[linux-yocto][PATCH 1/9] Revert "spi: zynq-qspi: use devm_platform_ioremap_resource() to simplify code"

From: Quanyang Wang This reverts commit ae91a439ac7ec8e1e925621293f45794de6366f2. Revert all mainline commits for spi-zynq-qspi.c and apply SDK patches to make sure that spi-zynq-qspi.c works in spi-master frame work. Signed-off-by: Quanyang Wang --- drivers/spi/spi-zynq-qspi.c | 4 +++- 1

Re: [linux-yocto] [V2] [linux-yocto-dev standard/xlnx-soc] New patches for v5.4

On 12/16/19 4:55 PM, Michal Simek wrote: On 16. 12. 19 9:53, Quanyang Wang wrote: On 12/16/19 4:46 PM, Michal Simek wrote: Hi, Naga: Can you please take a look at these spi patches? On 16. 12. 19 9:36, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce & Michal, Would you

Re: [linux-yocto][PATCH 1/9] Revert "spi: zynq-qspi: use devm_platform_ioremap_resource() to simplify code"

Hi Naga, On 12/16/19 7:40 PM, Naga Sureshkumar Relli wrote: Hi Quanyang, These patches are ok. but I didn't see any spi-nor.c or other frame work related patches to support dual parallel and stacked modes. Or am I missing something? These patches are in the git tree:

[linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

From: Quanyang Wang The zcu102 and zc706 boards support qspi dual mode, and in the dual mode, the erasesize for mtd device should be twice as much as the erasesize for each nor flash chip. Signed-off-by: Quanyang Wang --- drivers/mtd/spi-nor/spi-nor.c | 6 +- 1 file changed, 5

[linux-yocto] [linux-yocto-dev standard/xlnx-soc] mtd: spi-nor: fix erasesize for dual qspi mode

From: Quanyang Wang Hi Bruce & Michal, There is a fix for erasesize in dual mode in xilinx v4.19 kernel, but the spi-nor framework has changed in new version kernel, so I make this patch to fix it in new kernel. Would you please help review and merge these patches to linux-yocto-dev

[linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH] xlnx-soc: create v5.4/standard/xlnx-soc branch

Hi Bruce, Would you please help create branch v5.4/standard/xlnx-soc  on linux-yocto  and pull patches from linux-yocto-dev/v5.4/standard/xlnx-soc ? Thanks, Quanyang -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#8398):

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: SDK patches for kernel v5.5

Hi Michal, On 1/20/20 2:39 PM, Michal Simek wrote: On 18. 01. 20 0:53, Bruce Ashfield wrote: On Tue, Jan 14, 2020 at 9:38 PM Quanyang Wang wrote: Hi Bruce & Michal, Would you please help to remove the obsolete branch standard/xlnx-soc in linux-yocto-dev and apply these patches for v5.5 as

[linux-yocto] [linux-yocto-dev standard/xlnx-soc] usb: dwc3: gadget: fix compile error caused by an unclosed brace

From: Quanyang Wang When merging upstream commit da10bcdd6f70 ("usb: dwc3: gadget: Delay starting transfer"), a brace is lost. Add it to avoid compile error. Signed-off-by: Quanyang Wang --- drivers/usb/dwc3/gadget.c | 10 ++ 1 file changed, 6 insertions(+), 4 deletions(-) diff

Re: [V2] [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

Hi Bruce, On 2/6/20 5:42 AM, Bruce Ashfield wrote: In message: [V2] [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode on 06/01/2020 quanyang.w...@windriver.com wrote: From: Quanyang Wang The zcu102 and zc706 boards support qspi dual mode, and in the dual mode, the

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: SDK patches for kernel v5.5

these paches since they were for rc6. I will send a V2 SDK patches which contains tiny modification for context. Thanks, Quanyang Bruce Thanks, Quanyang On 1/20/20 2:53 PM, quanyang.wang wrote: Hi Michal, On 1/20/20 2:39 PM, Michal Simek wrote: On 18. 01. 20 0:53, Bruce Ashfield wrote

Re: [V2] [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

Hi Michal, On 2020/2/3 下午3:12, Michal Simek wrote: On 03. 02. 20 7:09, qwang2 wrote: Hi Amit, On 2020/2/3 下午1:51, Amit Kumar Mahapatra wrote: Hi, The patch looks fine to me. But I have few queries, @Michal Simek kindly share your thoughts. I had sent

Re: [V2] [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

/1] mtd: spi-nor: fix erasesize for dual qspi mode Ping. On 2020/1/6 下午9:33, quanyang.wang wrote: From: Quanyang Wang <mailto:quanyang.w...@windriver.com> The zcu102 and zc706 boards support qspi dual mode, and in the dual mode, the erasesize for mtd device should be

Re: [V2] [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

Ping. On 2020/1/6 下午9:33, quanyang.wang wrote: From: Quanyang Wang The zcu102 and zc706 boards support qspi dual mode, and in the dual mode, the erasesize for mtd device should be twice as much as the erasesize for each nor flash chip. Signed-off-by: Quanyang Wang --- V1 ---> V2: 1) use

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: SDK patches for kernel v5.5

Hi Bruce, Would you please help merge these patches? Thanks, Quanyang On 1/20/20 2:53 PM, quanyang.wang wrote: Hi Michal, On 1/20/20 2:39 PM, Michal Simek wrote: On 18. 01. 20 0:53, Bruce Ashfield wrote: On Tue, Jan 14, 2020 at 9:38 PM Quanyang Wang wrote: Hi Bruce & Michal, Would

[linux-yocto] [linux-yocto-dev standard/xlnx-soc] [PATCH 0/1] Fix ubifs test failed in v5.4

From: Quanyang Wang Hi all, This patch fixes an issue which occurrs in ubifs testcase. The steps to reproduce it is as below: root@xilinx-zynq:~# ubiformat /dev/mtd3 -y ubiformat: mtd3 (nor), size 6160384 bytes (5.8 MiB), 47 eraseblocks of 131072 bytes (128.0 KiB), min. I/O size 1 bytes

[linux-yocto][PATCH 1/1] xilinx-zynqmp: change CONFIG_REMOTEPROC from m to y

From: Quanyang Wang Since the upstream commit 6334150e9a36 ("remoteproc: don't allow modular build") has forbidden CONFIG_REMOTEPROC from being "m", so we change it to be "y". Or else there will be warning as below: -- CONFIG_REMOTEPROC - Config: CONFIG_REMOTEPROC From:

[linux-yocto][kernel-cache master] xilinx-zynqmp: change CONFIG_REMOTEPROC from m to y

From: Quanyang Wang Hi Bruce, Would you please help merge the patches to yocto-kernel-cache master branch? Thanks, Quanyang Quanyang Wang (1): xilinx-zynqmp: change CONFIG_REMOTEPROC from m to y bsp/xilinx-zynqmp/xilinx-zynqmp.cfg | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --

[V2] [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

From: Quanyang Wang The zcu102 and zc706 boards support qspi dual mode, and in the dual mode, the erasesize for mtd device should be twice as much as the erasesize for each nor flash chip. Signed-off-by: Quanyang Wang --- V1 ---> V2: 1) use nor->shift instead of 2. 2) put the implementation in

Re: [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

Hi Amit, On 2020/1/6 下午6:01, Amit Kumar Mahapatra wrote: Hi Quanyang, -Original Message- From: Naga Sureshkumar Relli Sent: Monday, January 6, 2020 2:59 PM To: Quanyang Wang ; Michal Simek ; Bruce ; Amit Kumar Mahapatra Cc: linux-yo...@yoctoproject.org Subject: RE:

Re: [linux-yocto][PATCH 1/1] mtd: spi-nor: fix erasesize for dual qspi mode

Hi Naga, On 12/18/19 5:59 PM, Michal Simek wrote: On 18. 12. 19 7:33, quanyang.w...@windriver.com wrote: From: Quanyang Wang The zcu102 and zc706 boards support qspi dual mode, and in the dual mode, the erasesize for mtd device should be twice as much as the erasesize for each nor flash

Re: [linux-yocto][PATCH 1/1] mtd: spi-nor: update ptr pointer after spi_nor_read_data

Hi Michal, On 1/6/20 3:16 PM, Michal Simek wrote: On 26. 12. 19 6:44, quanyang.w...@windriver.com wrote: From: Quanyang Wang In spi_nor_spimem_xfer_data, it will limit the transfer size according to nor->bouncebuf_size which is usually equal to PAGE_SIZE. This means that sometimes a

Re: [linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Michal, As below is the detail steps to reproduce this issue in SDK kernel: 1).  the aarch-none-linux-gnu-gcc v9.2 is here

[linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

From: Quanyang Wang When assigning value to a pointer as below: itapdly = (u32 [MMC_TIMING_MMC_HS400 + 1]) {0x0, 0x15, 0x15...0x0}; then using gcc 9.2.0 with -O2 to compile, the pointer itapdly will point to the stack of the function, and the values in this area is not initialized. The

[linux-yocto][linux-yocto v5.4/standard/xlnx-soc] sdhci: arasan: fix a bug caused by using yocto gcc v9.2.0

From: Quanyang Wang Hi Bruce & Michal, When using gcc 9.2.0 to build kernel, boot the kernel, any operation on SD card will occur the error: blk_update_request: I/O error, dev mmcblk0, sector 0 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0 Buffer I/O error on dev mmcblk0, logical block 0,

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] usb: dwc3: gadget: fix compile error caused by an unclosed brace

Hi Bruce, Would you please help merge this patch to linux-yocto-dev v5.5/standard/xlnx-soc? This branch also needs this patch. Thanks, Quanyang On 3/3/20 5:07 AM, Bruce Ashfield wrote: merged (this is specific to linux-yocto, so I've grabbed it) Cheers, Bruce On Fri, Feb 28, 2020 at

[linux-yocto][linux-yocto v5.4/standard/base][PATCH 1/1] fix memory leak when /proc/yaffs is read

From: Charles Manning commit 27f18203551940abf35826a66978daf1b8124c6b from git://www.aleph1.co.uk/yaffs2 Thanks to Jisheng Zhang for supplying this patch There is a kernel memory leak observed when the proc file /proc/yaffs is read. This reason is that in yaffs_proc_open, single_open is

[linux-yocto][kernel-cache master][PATCH 1/1] xilinx-zynqmp: add cpuidle support for zcu102

From: Quanyang Wang Because of upstream commit 788961462f347161 ("ARM: psci: cpuidle: Enable PSCI CPUidle driver"), the new kernel option CONFIG_ARM_PSCI_CPUIDLE is used to enable PSCI CPUidle driver. So enable CONFIG_ARM_PSCI_CPUIDLE for zcu102 board. Signed-off-by: Quanyang Wang ---

[linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 1/1] usb: dwc3: use PROPERTY_ENTRY_BOOL to initialize property entries

From: Quanyang Wang According to commit 5eb5afb07853d6e9 ("usb: dwc3: use proper initializers for property entries"), property entries should be initialized by proper initializer macros. If not, the elements "is_inline" and "length" in structure property_entry will be zeros, and the function

Re: [linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 1/1] usb: dwc3: use PROPERTY_ENTRY_BOOL to initialize property entries

On 4/3/20 8:41 AM, quanyang.wang wrote: From: Quanyang Wang According to commit 5eb5afb07853d6e9 ("usb: dwc3: use proper initializers for property entries"), property entries should be initialized by proper initializer macros. If not, the elements "is_inline" and &

Re: [linux-yocto][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Bruce, Would you please help merge this patch to linux-yocto-dev standard/xlnx-soc ? Thanks, Quanyang On 3/18/20 12:28 AM, quanyang.wang wrote: From: Quanyang Wang When assigning value to a pointer as below: itapdly = (u32 [MMC_TIMING_MMC_HS400 + 1]) {0x0, 0x15, 0x15...0x0

[V2][linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 0/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

From: Quanyang Wang V1 ---> V2: 1. modify commit description since the gcc with -O2 is not the root cause. 2. fix some format defects. Quanyang Wang (1): sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Re: [linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Michal, Is the xilinx kernel 5.4 finished? I check the git tree https://github.com/Xilinx/linux-xlnx.git and  there is no xlnx_rebase_v5.4 branch. Thanks, Quanyang On 2020/3/17 下午11:25, Michal Simek wrote: Hi, tbh I think that it will be good to start to think how to switch 5.4 to be

Re: [linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Michal, It's strange that this issue only exists with SDK 4.19 patches for sdhci-of-arasan.c and gcc 9.2. Once the code is changed, this issue disappears. It means that the over optimization of gcc 9.2 is not the root cause. Since this issue breaks down the mmc controller in linux-yocto 5.4

Re: [linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Michal, On 2020/3/18 下午12:40, quanyang.wang wrote: Hi Michal, On 2020/3/17 下午5:24, Michal Simek wrote: Hi, I have tested the latest Xilinx kernel based on v5.4 at github.com/Xilinx/linux-xlnx and I built it by toolchain you pointed me to and I can't see any issue. I find

Re: [linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Michal, On 2020/3/17 下午5:24, Michal Simek wrote: Hi, I have tested the latest Xilinx kernel based on v5.4 at github.com/Xilinx/linux-xlnx and I built it by toolchain you pointed me to and I can't see any issue. I find that the reason why this issue only exists in SDK v4.19 but not in SDK

Re: [linux-yocto][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

Hi Bruce, On 3/20/20 6:39 AM, Bruce Ashfield wrote: On Tue, Mar 17, 2020 at 12:29 PM wrote: From: Quanyang Wang When assigning value to a pointer as below: itapdly = (u32 [MMC_TIMING_MMC_HS400 + 1]) {0x0, 0x15, 0x15...0x0}; then using gcc 9.2.0 to compile, in some condition the

[linux-yocto][kernel-cache master][PATCH 1/1] bsp: delete obsolete kernel option CONFIG_MTD_M25P80

From: Quanyang Wang Because of upstream commit b35b9a10362d ("mtd: spi-nor: Move m25p80 code in spi-nor.c"), CONFIG_MTD_M25P80 is not used anymore. Signed-off-by: Quanyang Wang --- bsp/marvell-cn96xx/marvell-cn96xx.cfg | 1 - bsp/nxp-ls20xx/nxp-ls20xx.cfg | 1 -

Re: [linux-yocto][PATCH 1/1] sdhci: arasan: use an array of integers for *_itap/otap_delays instead of macro *_ITAP/OTAP_DELAYS

On 2020/3/20 下午4:03, quanyang.wang wrote: Hi Bruce, On 3/20/20 6:39 AM, Bruce Ashfield wrote: On Tue, Mar 17, 2020 at 12:29 PM wrote: From: Quanyang Wang When assigning value to a pointer as below: itapdly = (u32 [MMC_TIMING_MMC_HS400 + 1]) {0x0, 0x15, 0x15...0x0}; then using

[linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 1/1] Revert "usb: gadget: configfs: Add support for setting maximum supported speed"

From: Quanyang Wang The mainline commit a02497033e8e ("usb: gadget: configfs: Add max_speed setting") has added the same functions and conflicts with this SDK patch. So revert this SDK patch to avoid building error. Signed-off-by: Quanyang Wang --- drivers/usb/gadget/configfs.c | 45

Re: [linux-yocto][kernel-cache master][PATCH 1/1] xilinx-zynqmp: add cpuidle support for zcu102

Hi Bruce, Would you please help merge this patch to kernel-cache yocto-5.4 branch? Thanks, Quanyang On 3/17/20 10:41 AM, Bruce Ashfield wrote: merged. Bruce In message: [linux-yocto][kernel-cache master][PATCH 1/1] xilinx-zynqmp: add cpuidle support for zcu102 on 16/03/2020

Re: [linux-yocto][kernel-cache master][PATCH 1/1] bsp: delete obsolete kernel option CONFIG_MTD_M25P80

Hi Bruce, Would you please help merge this patch to kernel-cache yocto-5.4 branch? Thanks, Quanyang On 3/22/20 9:30 AM, Bruce Ashfield wrote: merged. Bruce In message: [linux-yocto][kernel-cache master][PATCH 1/1] bsp: delete obsolete kernel option CONFIG_MTD_M25P80 on 20/03/2020

[linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 1/1] PCI: XDMA PL: use pci_parse_request_of_pci_ranges instead of devm_of_pci_get_host_bridge_resources

From: Quanyang Wang Because of mainline commit 3b55809cf91f ("PCI: Make devm_of_pci_get_host_bridge_resources() static"), use pci_parse_request_of_pci_ranges instead of devm_of_pci_get_host_bridge_resources(). Signed-off-by: Quanyang Wang --- drivers/pci/controller/pcie-xdma-pl.c | 17

[linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 0/1] xilinx-zynqmp: fix a porting issue

From: Quanyang Wang Hi Bruce, Would you please help merge this patch to linux-yocto-dev standard/xlnx-soc branch? This patch just fix a porting issue that the function devm_of_pci_get_host_bridge_resources isn't global anymore in v5.7 and we should use new function

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 0/2] linux-yocto-dev-v5.7: fix usb issue

From: Quanyang Wang Hi Bruce & Michal, Would you please help review these 2 patches and merge them to linux-yocto-dev standard/xlnx-soc? These 2 patches fixes porting issues, not SDK bug. Thanks, Quanyang Quanyang Wang (2): usb: dwc3: use PROPERTY_ENTRY_BOOL to initialize property entries

[linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 2/2] firmware: xilinx: make pm_ioctl_id identical to SDK definitions

From: Quanyang Wang The enum pm_ioctl_id should be identical to the definitions in SDK, or else according to the value of pm_ioctl_id, secure monitor will execute unexpected instructions. Signed-off-by: Quanyang Wang --- include/linux/firmware/xlnx-zynqmp.h | 4 ++-- 1 file changed, 2

[linux-yocto] [kernel-cache master][PATCH 1/1] xilinx-zynqmp: enable CONFIG_RAS explicitly

From: Quanyang Wang Because of the mainline commit 8c8ff55b4da7 ("PCI/AER: Don't select CONFIG_PCIEAER by default"), CONFIG_PCIEAER will not be enabled by default, this results that CONFIG_RAS which is seleted by CONFIG_PCIEAER will not be enabled either. Since CONFIG_EDAC_ZYNQMP_OCM depends on

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 1/2] PCI: XDMA PL: use pci_parse_request_of_pci_ranges instead of devm_of_pci_get_host_bridge_resources

From: Quanyang Wang Because of mainline commit 3b55809cf91f ("PCI: Make devm_of_pci_get_host_bridge_resources() static"), use pci_parse_request_of_pci_ranges instead of devm_of_pci_get_host_bridge_resources(). Signed-off-by: Quanyang Wang Signed-off-by: Bruce Ashfield ---

[linux-yocto][linux-yocto-dev standard/xlnx-soc][PATCH 1/1] arm: zynq: delete AFLAGS_suspend.o to fix compile warning

From: Quanyang Wang When using arm-linux-gnueabihf-gcc v9.2.0 to compile kernel, there will be warning info as below: cc1: warning: switch '-mcpu=cortex-a9' conflicts with '-march=armv7-a' switch This is because that if "-march" and "-mcpu" options are passed to gcc together, gcc will check if

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 0/2] Delete mainline content

From: Quanyang Wang Hi Bruce, This 2 patches are to delete the redundant content which is introduced by mainline. Would you please help merge them to standard/xlnx-soc ? Thanks, Quanyang Quanyang Wang (2): drm: xlnx: delete mainline Kconfig content arm64: zynqmp: delete mainline dts

[linux-yocto][PATCH 2/2] arm64: zynqmp: delete mainline dts content

From: Quanyang Wang This is to delete the unnecessary content which is introduced from mainline. Signed-off-by: Quanyang Wang --- arch/arm64/boot/dts/xilinx/zynqmp.dtsi | 21 - 1 file changed, 21 deletions(-) diff --git a/arch/arm64/boot/dts/xilinx/zynqmp.dtsi

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH] PCI: XDMA PL: use pci_parse_request_of_pci_ranges instead of devm_of_pci_get_host_bridge_resources

From: Quanyang Wang Because of mainline commit 3b55809cf91f ("PCI: Make devm_of_pci_get_host_bridge_resources() static"), use pci_parse_request_of_pci_ranges instead of devm_of_pci_get_host_bridge_resources(). Signed-off-by: Quanyang Wang Signed-off-by: Bruce Ashfield ---

Re: [linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH] net: macb: drop wol modifications from upstream

Hi Bruce, On 8/4/20 9:50 PM, Bruce Ashfield wrote: I've taken care of this as part of merging paulg's latest 5.2 -stable updates. This keeps coming back .. and I've made a note to look out for it on future merges. It seems that this patch still hasn't been merged to linux-yocto

[linux-yocto][linux-yocto v5.4/standard/xlnx-soc][PATCH] net: macb: drop wol modifications from upstream

From: Quanyang Wang The kernel applies the mainline commit ced4799d0637 ("net: macb: mark device wake capable when magic-packet property present") on the magic packet WOL feature of macb. But in the SDK commit 3478668d12f0 ("net: macb: Add PM support"), the magic packet WOL support is removed

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH] net: macb: drop wol modifications from upstream

From: Quanyang Wang The v5.8-rc5 kernel applies the commit 253fe0943504 ("net: macb: fix macb_get/set_wol() when moving to phylink") and the commit ced4799d0637 ("net: macb: mark device wake capable when magic-packet property present") on the magic packet WOL feature of macb. But in the SDK

[linux-yocto] [linux-yocto v5.4/standard/xlnx-soc][PATCH] usb: dwc3: gadget: declare __dwc3_gadget_kick_transfer as global

From: Quanyang Wang Since the function "__dwc3_gadget_kick_transfer" is used by drivers/usb/dwc3/gadget_hibernation.c, so declare it as global. This is to fix a compile error as below: drivers/usb/dwc3/gadget.c:1248:12: error: static declaration of '__dwc3_gadget_kick_transfer' follows

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH] net: macb: drop wol modifications from upstream

Hi Bruce, On 7/15/20 10:24 PM, Bruce Ashfield wrote: In message: [linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH] net: macb: drop wol modifications from upstream on 15/07/2020 quanyang.w...@windriver.com wrote: From: Quanyang Wang The v5.8-rc5 kernel applies the commit 253fe0943504

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 0/1] export the maximum and minimum voltage and temperature to userspace

Hi Michal, On 7/28/20 1:20 PM, Michal Simek wrote: On 27. 07. 20 10:53, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce and Michal, One of our customers want to use more xadc sysfs interfaces, so export the maximum and minimum voltage and temperature to userspace. Would

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 0/1] export the maximum and minimum voltage and temperature to userspace

From: Quanyang Wang Hi Bruce and Michal, One of our customers want to use more xadc sysfs interfaces, so export the maximum and minimum voltage and temperature to userspace. Would you please help review and merge this patch to linux-yocto-dev standard/xlnx-soc? Thanks, Quanyang Chen.Liu (1):

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH 1/1] xilinx-xadc: export the maximum and minimum voltage and temperature to userspace

From: "Chen.Liu" Export more xadc sysfs interfaces to userspace. Signed-off-by: Chen.Liu [Quanyang: adapted the patch to solve context issues and add some commit descriptions] Signed-off-by: Quanyang Wang --- drivers/iio/adc/xilinx-xadc-core.c | 17 ++--- 1 file changed, 14

[linux-yocto] [linux-yocto-dev standard/xlnx-soc][PATCH] usb: dwc3: use PROPERTY_ENTRY_BOOL to initialize property entries

From: Quanyang Wang According to commit 5eb5afb07853d6e9 ("usb: dwc3: use proper initializers for property entries"), property entries should be initialized by proper initializer macros. If not, the elements "is_inline" and "length" in structure property_entry will be zeros, and the function

[linux-yocto] [kernel-cache master][PATCH] bsp: enable CONFIG_MEDIA_PLATFORM_SUPPORT kernel option

From: Quanyang Wang Because of the commit 06b93644f4d1 ("media: Kconfig: add an option to filter in/out platform drivers"), CONFIG_MEDIA_PLATFORM_SUPPORT must be enabled when we want to enable CONFIG_V4L_PLATFORM_DRIVERS. Signed-off-by: Quanyang Wang --- bsp/xilinx-zynq/xilinx-zynq.cfg |

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: porting SDK patches to kernel v5.10

Hi Michal, On 11/26/20 8:20 PM, Michal Simek wrote: Hi, On 26. 11. 20 11:19, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce, Would you please help remove the obsolete branch standard/xlnx-soc in linux-yocto-dev and apply these patches for v5.10 as below to the new branch

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: porting SDK patches to kernel v5.10

Hi Michal, On 11/27/20 6:28 PM, Michal Simek wrote: Hi, On 27. 11. 20 11:22, quanyang.wang wrote: Hi Michal, On 11/26/20 8:20 PM, Michal Simek wrote: Hi, On 26. 11. 20 11:19, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce, Would you please help remove the obsolete

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: porting SDK patches to kernel v5.10

Hi Michal, On 11/26/20 9:10 PM, Michal Simek wrote: On 26. 11. 20 14:06, quanyang.wang wrote: Hi Michal, On 11/26/20 8:20 PM, Michal Simek wrote: Hi, On 26. 11. 20 11:19, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce, Would you please help remove the obsolete branch

[linux-yocto] [PATCH 1/4] spi: spi-zynqmp-gqspi: transmit dummy circles by using controller's internal functionality

From: Quanyang Wang When reading from mt25qu512a on zcu102 board, it will return wrong data. There are two problems: 1. The gqspi controller has a functionality to send dummy clock circles. When write data with the fields [receive, transmit, data_xfer] = [0,0,1] to the Generic FIFO, the

[linux-yocto] [PATCH 2/4] spi: spi-zynqmp-gqspi: add mutex locking for exec_op

From: Quanyang Wang The spi-mem framework has no locking to prevent ctlr->mem_ops->exec_op from concurrency. So add the locking to zynqmp_qspi_exec_op. Signed-off-by: Quanyang Wang --- drivers/spi/spi-zynqmp-gqspi.c | 5 + 1 file changed, 5 insertions(+) diff --git

[linux-yocto] [PATCH 3/4] spi: spi-zynqmp-gqspi: use wait_for_completion_timeout instead of wait_for_completion_interruptible_timeout

From: Quanyang Wang The function wait_for_completion_interruptible_timeout will return a non-zero value -ERESTARTSYS when the current process receives a signal "SIGKILL". So there is a scene that the function wait_for_completion_interruptible_timeout exits not because the handler zynqmp_qspi_irq

[linux-yocto] [PATCH 4/4] spi: spi-zynqmp-gqspi: fix zynqmp_qspi_read_op assign wrong transfer_len

From: Quanyang Wang For some scenarios, the wrong transfer_len is written to genfifoentry. Take 10 bytes data reading operation for instance, 8 bytes will be read from the flash by using DMA mode and 2 bytes left in IO mode. Thus we need to create an 8-byte size transfer genfifoentry in

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: porting SDK patches to kernel v5.10

Hi Michal, On 11/26/20 8:20 PM, Michal Simek wrote: Hi, On 26. 11. 20 11:19, quanyang.w...@windriver.com wrote: From: Quanyang Wang Hi Bruce, Would you please help remove the obsolete branch standard/xlnx-soc in linux-yocto-dev and apply these patches for v5.10 as below to the new branch

Re: [linux-yocto] [linux-yocto-dev standard/xlnx-soc] xlnx-soc: porting SDK patches to kernel v5.10

Hi Amit, On 12/3/20 6:36 PM, Amit Kumar Mahapatra wrote: Hello Quanyang, Thank you for the patches. With 0003-spi-spi-zynqmp-gqspi-use-wait_for_completion_timeout.patch I am seeing mtd_debug erase, write and read failures. root@xilinx-zc1751-dc1-2020_2:~# cat /proc/mtd dev:size

[linux-yocto] [PATCH 2/3] drm: xlnx: configure alpha value to make graphic layer opaque

From: Quanyang Wang Since graphics layer is primary, and video layer is overaly, we need to configure the V_BLEND_SET_GLOBAL_ALPHA_REG register to make graphic layer opaque by default, or else graphic layer will be transparent and invisible. Signed-off-by: Quanyang Wang ---

[linux-yocto] [PATCH 1/3] drm: xlnx: set PLL_MODE_FRAC mode to VPLL_FRAC_CFG by re-enabling disp->pclk

From: Quanyang Wang When the function clk_set_rate configures the rate of disp->pclk, zynqmp_pm_set_pll_frac_mode will be called to set VPLL's mode to be PLL_MODE_FRAC or PLL_MODE_INT by invoking an SMC call to ATF. But in ATF, the service pm_clock_set_pll_mode doesn't really set the

[linux-yocto] [PATCH 3/3] drm/xlnx: change drm_driver name from zynqmp-dpsub to xlnx

From: Quanyang Wang When booting an X11 graphic system, Xorg will load armsoc_drv.so as X.Org Video Driver. In armsoc_drv.so, "xlnx" is the driver_name of struct drmmode_interface, and it must match the name used in the kernel driver. So change the drm_driver name to "xlnx" to maintain

  1   2   3   4   5   >