LIBUSB_ERROR_TIMEOUT with big rootfs.tar.bz2

[jessie80]

Hi customer met LIBUSB_ERR0R_TIMEOUT during downloading the roofs.tar.bz2 which is compiled with bitbake fsl-image-validation-imx. but smaller rootfs.tar.bz is working well.

below is the uuu log..
C:\uuu_cg>uuu -v -V lgecg_emmc_download.uuu
uuu (Universal Update Utility) for nxp imx chips -- libuuu_1.2.135-0-gacaf035

Build in config:
Pctl Chip Vid Pid BcdVersion
==================================================
SDPS: MX8QXP 0x1fc9 0x012f [0x0002..0xffff]
SDPS: MX8QM 0x1fc9 0x0129 [0x0002..0xffff]
SDPS: MX28 0x15a2 0x004f
SDPS: MX815 0x1fc9 0x013e
SDP: MX7D 0x15a2 0x0076
SDP: MX6Q 0x15a2 0x0054
SDP: MX6D 0x15a2 0x0061
SDP: MX6SL 0x15a2 0x0063
SDP: MX6SX 0x15a2 0x0071
SDP: MX6UL 0x15a2 0x007d
SDP: MX6ULL 0x15a2 0x0080
SDP: MX6SLL 0x1fc9 0x0128
SDP: MX7ULP 0x1fc9 0x0126
SDP: MXRT106X 0x1fc9 0x0135
SDP: MX8MM 0x1fc9 0x0134
SDP: MX8MQ 0x1fc9 0x012b
SDPU: SPL 0x0525 0xb4a4 [0x0000..0x04ff]
SDPV: SPL1 0x0525 0xb4a4 [0x0500..0x9998]
SDPU: SPL 0x0525 0xb4a4 [0x9999..0x9999]
FBK: 0x066f 0x9afe
FBK: 0x066f 0x9bff
FB: 0x0525 0xa4a5
FB: 0x18d1 0x0d02
Wait for Known USB Device Appear...
New USB Device Attached at 1:13
1:13>Start Cmd:SDP: boot -f u-boot.imx
6400%libusb: error [windows_transfer_callback] detected I/O error 1167: [1167] 장치가 연결되지 않았습니다.
libusb: error [hid_submit_bulk_transfer] HID transfer failed: [1167] 장치가 연결되지 않았습니다.
1:13>Okay
New USB Device Attached at 1:13
1:13>Start Cmd:FB: ucmd setenv fastboot_dev mmc
1:13>Okay
1:13>Start Cmd:FB: ucmd setenv mmcdev ${emmc_dev}
1:13>Okay
1:13>Start Cmd:FB: flash bootloader u-boot.imx
0x400000000x2001:13>Okay
1:13>Start Cmd:FB: ucmd if env exists emmc_ack; then ; else setenv emmc_ack 0; fi;
1:13>Okay
1:13>Start Cmd:FB: ucmd setenv emmc_cmd mmc partconf ${emmc_dev} ${emmc_ack} 1 0
1:13>Okay
1:13>Start Cmd:FB: ucmd if test "${emmc_skip_fb}" != "yes"; then run emmc_cmd; fi
1:13>Okay
1:13>Start Cmd:FB: ucmd setenv emmc_cmd mmc bootbus ${emmc_dev} 2 2 1;
1:13>Okay
1:13>Start Cmd:FB: ucmd if test "${emmc_skip_fb}" != "yes"; then run emmc_cmd; fi
1:13>Okay
1:13>Start Cmd:FB: ucmd echo '[warkap] step #0 before download -f zImage'
1:13>Okay
1:13>Start Cmd:FB: ucmd setenv fastboot_buffer ${loadaddr}
1:13>Okay
1:13>Start Cmd:FB: download -f zImage
1:13>Okay
1:13>Start Cmd:FB: ucmd echo '[warkap] step #1 after download -f zImage'
1:13>Okay
1:13>Start Cmd:FB: ucmd setenv fastboot_buffer ${fdt_addr}
1:13>Okay
1:13>Start Cmd:FB: download -f imx6sx-sdb-lgecg.dtb
1:13>Okay
1:13>Start Cmd:FB: ucmd setenv fastboot_buffer ${initrd_addr}
1:13>Okay
1:13>Start Cmd:FB: download -f _initramfs.cpio.gz.uboot
1:13>Okay
1:13>Start Cmd:FB: acmd ${kboot} ${loadaddr} ${initrd_addr} ${fdt_addr}
1:13>Okay
New USB Device Attached at 1:13
1:13>Start Cmd:FBK: ucmd while [ ! -e /dev/mmcblkboot0 ]; do sleep 1; echo "wait for /dev/mmcblkboot* appear"; done;
1:13>Okay
1:13>Start Cmd:FBK: ucmd dev=ls /dev/mmcblk*boot*; dev=($dev); dev=${dev[0]}; dev=${dev#/dev/mmcblk}; dev=${dev%boot*}; echo $dev > /tmp/mmcdev;
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; PARTSTR=$'10M,500M,0c\n600M,,83\n'; echo "$PARTSTR" | sfdisk --force /dev/mmcblk${mmc}
Checking that no-one is using this disk right now ... OK

Disk /dev/mmcblk3: 3.7 GiB, 3959422976 bytes, 7733248 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0x67e6b631

Old situation:

Device Boot Start End Sectors Size Id Type
/dev/mmcblk3p1 20480 1044479 1024000 500M c W95 FAT32 (LBA)
/dev/mmcblk3p2 1228800 7733247 6504448 3.1G 83 Linux

Created a new DOS disklabel with disk identifier 0xe090e6d2.
/dev/mmcblk3p1: Created a new partition 1 of type 'W95 FAT32 (LBA)' and of size 500 MiB.
/dev/mmcblk3p2: Created a new partition 2 of type 'Linux' and of size 3.1 GiB.
/dev/mmcblk3p3: Done.

New situation:
Disklabel type: dos
Disk identifier: 0xe090e6d2

Device Boot Start End Sectors Size Id Type
/dev/mmcblk3p1 20480 1044479 1024000 500M c W95 FAT32 (LBA)
/dev/mmcblk3p2 1228800 7733247 6504448 3.1G 83 Linux

The partition table has been altered.
Calling ioctl() to re-read partition table.
Syncing disks.
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; dd if=/dev/zero of=/dev/mmcblk${mmc} bs=1k seek=4096 count=1
1:13>Okay
1:13>Start Cmd:FBK: ucmd sync
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; while [ ! -e /dev/mmcblk${mmc}p1 ]; do sleep 1; done
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mkfs.vfat /dev/mmcblk${mmc}p1
mkfs.fat 4.1 (2017-01-24)
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mkdir -p /mnt/fat
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mount -t vfat /dev/mmcblk${mmc}p1 /mnt/fat
1:13>Okay
1:13>Start Cmd:FBK: ucp zImage t:/mnt/fat
99%1:13>Okay
1:13>Start Cmd:FBK: ucp imx6sx-sdb-lgecg.dtb t:/mnt/fat
152%1:13>Okay
1:13>Start Cmd:FBK: ucmd umount /mnt/fat
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mkfs.ext3 -F -E nodiscard /dev/mmcblk${mmc}p2
Creating filesystem with 813056 4k blocks and 203600 inodes
Filesystem UUID: ea45f09c-b6ed-464c-96e0-7286d9e0c1ba
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912

Allocating group tables: done
Writing inode tables: done
Creating journal (16384 blocks): done
Writing superblocks and filesystem accounting information: done

1:13>Okay
1:13>Start Cmd:FBK: ucmd mkdir -p /mnt/ext3
1:13>Okay
1:13>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mount /dev/mmcblk${mmc}p2 /mnt/ext3
1:13>Okay
1:13>Start Cmd:FBK: acmd export EXTRACT_UNSAFE_SYMLINKS=1; tar -jx -C /mnt/ext3
1:13>Okay
1:13>Start Cmd:FBK: ucp imx6sxsabresd.rootfs.tar.bz2 t:-
48%1:13>Fail Bulk(W):LIBUSB_ERROR_TIMEOUT

Could you please check and help me to find the root cause?

[nxpfrankli]

I may cause slow sd\emmc write speed.

Can you

https://github.com/NXPmicro/imx-uuc/blob/master/ufb.c

line 532:
ret = write_file(g_open_file, p, size);
if (ret < 0)
key = FAIL;

you can print out how much time consumed by this function

[nxpfrankli]

Do you have uart log?

[jessie80]

for your first comment, it seems it is not uuu source code itself. If there is similar code part on uuu (window version) to check the write speed, i'll try with customer. Could you please guide me?
Or i've never seen this imx-uuc before. So, how could i modify and apply to system regarding of ufb.c, i'll try..

Anyway, below is the uart log.
U-Boot 2018.03-00044-ga3070e6-dirty (Aug 30 2019 - 11:22:41 +0900)

CPU:   Freescale i.MX6SX rev1.2 996 MHz (running at 792 MHz)
CPU:   Extended Commercial temperature grade (-20C to 105C) at 48C
Reset cause: POR
Model: Freescale i.MX6 SoloX SDB RevB Board
Board: MX6SX SABRE SDB rev@
DRAM:  256 MiB
PMIC: PFUZE100! DEV_ID=0x11 REV_ID=0x21
MMC:   FSL_SDHC: 0, FSL_SDHC: 1
Loading Environment from MMC... MMC Device 2 not found
*** Warning - No MMC card found, using default environment

Failed (-5)
Display: Hannstar-XGA (1024x768)
Video: 1024x768x18
In:    serial
Out:   serial
Err:   serial
MMC Device 2 not found
no mmc device at slot 2
flash target is MMC:2
MMC Device 2 not found
invalid mmc device 2
Net:
Error: ethernet@02188000 address not set.
No ethernet found.
Fastboot: Normal
Boot from USB for mfgtools
Use default environment for                              mfgtools
Run bootcmd_mfg: run mfgtool_args;if iminfo ${initrd_addr}; then if test ${tee} = yes; then bootm ${tee_addr} ${initrd_addr} ${fdt_addr}; else bootz ${loadaddr} ${initrd_addr} ${fdt_addr}; fi; else echo "Run fastboot ..."; fastboot 0; fi;
Hit any key to stop autoboot:  0

Checking Image at 86800000 ...

Unknown image format!
Run fastboot ...
flash target is MMC:1
Starting download of 629760 bytes
....
downloading of 629760 bytes finished
writing to partition 'bootloader'
support sparse flash partition for bootloader
Initializing 'bootloader'
switch to partitions #1, OK
mmc1(part 1) is current device
Writing 'bootloader'

MMC write: dev # 1, block # 2, count 1230 ... 1230 blocks written: OK
Writing 'bootloader' DONE!
[warkap] step #0 before download -f zImage
Starting download of 6870616 bytes
....................................................
downloading of 6870616 bytes finished
[warkap] step #1 after download -f zImage
Starting download of 42895 bytes

downloading of 42895 bytes finished
Starting download of 12348071 bytes
..........................................................................
....................
downloading of 12348071 bytes finished
Kernel image @ 0x80800000 [ 0x000000 - 0x68d658 ]

Loading init Ramdisk from Legacy Image at 86800000 ...

   Image Name:   fsl-image-mfgtool-initramfs-imx6
   Image Type:   ARM Linux RAMDisk Image (uncompressed)
   Data Size:    12348007 Bytes = 11.8 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK

Flattened Device Tree blob at 83000000

   Booting using the fdt blob at 0x83000000
   Using Device Tree in place at 83000000, end 8300d78e
switch to ldo_bypass mode!

Starting kernel ...

vddarm: bypassed regulator has no supply!
vddarm: failed to get the current voltage(-517)
vddsoc: bypassed regulator has no supply!
vddsoc: failed to get the current voltage(-517)
imx6sx-pinctrl 20e0000.iomuxc: Invalid fsl,pins or pins property in node /soc/aips-bus@02000000/iomuxc@020e0000/imx6x-sdb/wdoggrp
gpio-9 (cts): _gpiod_direction_output_raw: tried to set a GPIO tied to an IRQ as output
gpio-12 (cts): _gpiod_direction_output_raw: tried to set a GPIO tied to an IRQ as output
fec 2188000.ethernet (unnamed net_device) (uninitialized): Invalid MAC address: 00:00:00:00:00:00
caam-snvs 20cc000.caam-snvs: can't get snvs clock
Found New UDC: ci_hdrc.0
ci_hdrc.0 0
ffs.utp0
run utp at /dev/usb-utp0/ep0
.
uuu fastboot client 1.0.0 [built Apr 14 2019 13:43:48]
Start init usb
write string
Start handle command
uuc /dev/utp
uuc 0.5 [built Apr 14 2019 13:43:48]
UTP: Waiting for /dev/utp to appear
run shell cmd: while [ ! -e /dev/mmcblkboot0 ]; do sleep 1; echo "wait for /dev/mmcblkboot* appear"; done;
run shell cmd: dev=ls /dev/mmcblk*boot*; dev=($dev); dev=${dev[0]}; dev=${dev#/dev/mmcblk}; dev=${dev%boot*}; echo $dev > /tmp/mmcdev;
run shell cmd: mmc=cat /tmp/mmcdev; PARTSTR=$'10M,500M,0c\n600M,,83\n'; echo "$PARTSTR" | sfdisk --force /dev/mmcblk${mmc}
Partition #1 contains a vfat signature.
Partition #2 contains a ext3 signature.
run shell cmd: mmc=cat /tmp/mmcdev; dd if=/dev/zero of=/dev/mmcblk${mmc} bs=1k seek=4096 count=1
1+0 records in
1+0 records out
1024 bytes (1.0 kB, 1.0 KiB) copied, 0.00171967 s, 595 kB/s
run shell cmd: sync
run shell cmd: mmc=cat /tmp/mmcdev; while [ ! -e /dev/mmcblk${mmc}p1 ]; do sleep 1; done
run shell cmd: mmc=cat /tmp/mmcdev; mkfs.vfat /dev/mmcblk${mmc}p1
run shell cmd: mmc=cat /tmp/mmcdev; mkdir -p /mnt/fat
run shell cmd: mmc=cat /tmp/mmcdev; mount -t vfat /dev/mmcblk${mmc}p1 /mnt/fat
WOpen:/mnt/fat
WOpen:/mnt/fat/zImage
WOpen:/mnt/fat
WOpen:/mnt/fat/imx6sx-sdb-lgecg.dtb
run shell cmd: umount /mnt/fat
run shell cmd: mmc=cat /tmp/mmcdev; mkfs.ext3 -F -E nodiscard /dev/mmcblk${mmc}p2
mke2fs 1.43.8 (1-Jan-2018)
run shell cmd: mkdir -p /mnt/ext3
run shell cmd: mmc=cat /tmp/mmcdev; mount /dev/mmcblk${mmc}p2 /mnt/ext3
run shell cmd: export EXTRACT_UNSAFE_SYMLINKS=1; tar -jx -C /mnt/ext3
WOpen:-

[guwhite]

I'm having the same issue and with a similar serial log. Here are a few levers I've tried (Remove # to invoke).

#Levers for available free memory space
#FBK: ucmd echo 'vm.min_free_kbytes = 65536' > /etc/sysctl.d/01_minfreekbytes.conf
#FBK: ucmd echo 'vm.swappiness = 80' > /etc/sysctl.d/swappiness.conf
#FBK: ucmd sysctl -p /etc/sysctl.d/01_minfreekbytes.conf
#FBK: ucmd sysctl -p /etc/sysctl.d/swappiness.conf

#Levers for CPU clocking
#FBK: ucmd echo powersave > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor

#Levers for CPU enable
#FBK: ucmd echo "0" >/sys/devices/system/cpu/cpu1/online
#FBK: ucmd echo "0" >/sys/devices/system/cpu/cpu2/online
#FBK: ucmd echo "0" >/sys/devices/system/cpu/cpu3/online

#Levers for disk caching
#FBK: ucmd echo 100 > /proc/sys/vm/dirty_expire_centisecs
#FBK: ucmd echo 100 > /proc/sys/vm/dirty_writeback_centisecs

The only thing in the above that seems to affect the results is to limit the CPU to only cpu0 (it takes a long time but it passes). Turning on the other CPUs will eventually fail but at different places. In my scenario, enabling CPU0 and CPU3 will result in the script failing at the "unmount" command for some reason. Looking for suggestions.

[nxpfrankli]

Does #137 can work around this problem?

It's strange problem. Only 40MB used by ufb. I am not sure why usb can't allocate 64k memory. Our USB driver owner is checking it.

[guwhite]

#137 does not help in my scenario. Our scripts are different from above by first copying the bz2 to a temporary directory and untaring from there to disk instead of directly host copy / untar to disk.

[jessie80]

@nxpfrankli
i'm waiting the feedback from customer. as i know, he is on the business trip. he may be back in next Monday. I'll update soon.

[jessie80]

@nxpfrankli
According to the customer , with your temporal binary, TIMEOUT issue is resolved.
Is there any progress from USBdriver owner?

[nxpfrankli]

not yet. Not sure why there are not continue 64K memory at that time

[nxpfrankli]

5.4 kernel's usb support scatter list. Q2 release will include this change.
this problem can be fixed at 5.4 kernel.

[franklin-woo]

I got a same issue.
My rootfs file size 674mb.
I built the yocto(zeus) and kernel version is 5.4.3.

Below log is some part of uuu verbose log.

Allocating group tables: done
Writing inode tables: done
Creating journal (8192 blocks): done
Writing superblocks and filesystem accounting information: done

1:1>Okay (2.124s)
1:1>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mkfs.ext4 /dev/mmcblk${mmc}p4
Discarding device blocks: done
Creating filesystem with 841680 4k blocks and 210496 inodes
Filesystem UUID: d3b6b55d-1798-4021-9418-b56e3c7e5ff4
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200

Allocating group tables: done
Writing inode tables: done
Creating journal (16384 blocks): done
Writing superblocks and filesystem accounting information: done

1:1>Okay (3.781s)
1:1>Start Cmd:FBK: ucmd mkdir -p /mnt/rootfs
1:1>Okay (0.027s)
1:1>Start Cmd:FBK: ucmd mmc=cat /tmp/mmcdev; mount -t ext4 /dev/mmcblk${mmc}p2 /mnt/rootfs
1:1>Okay (0.105s)
1:1>Start Cmd:FBK: acmd tar -jx -C /mnt/rootfs
1:1>Okay (0.051s)
1:1>Start Cmd:FBK: ucp firmware/fsl-image-qt5-rootfs.tar.bz2 t:-
4%1:1>Fail Bulk(R):LIBUSB_ERROR_NO_DEVICE(165.7s)

And the kernel log.

U-Boot 2019.04-lf-5.4.y_v2019.04+g228843cdf5 (Jun 24 2020 - 19:00:51 +0900)

CPU: Freescale i.MX6SOLO rev1.3 996 MHz (running at 792 MHz)
CPU: Extended Commercial temperature grade (-20C to 105C) at 45C
Reset cause: POR
Model: Freescale i.MX6 DualLite SABRE Smart Device Board
Board: MX6-SabreSD
I2C: ready
DRAM: 512 MiB

........

Starting kernel ...

[ 0.000000] Booting Linux on physical CPU 0x0
[ 0.000000] Linux version 5.4.3+gf8118585ee3c (frank@frank-desktop) (gcc version 9.2.0 (GCC)) #1 SMP PREEMPT Thu Jun 25 17:42:03 KST 2020
[ 0.000000] CPU: ARMv7 Processor [412fc09a] revision 10 (ARMv7), cr=10c5387d
[ 0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing instruction cache
[ 0.000000] OF: fdt: Machine model: Freescale i.MX6 DualLite SABRE Smart Device Board
[ 0.000000] Memory policy: Data cache writeback
[ 0.000000] Reserved memory: created CMA memory pool at 0x1c000000, size 320 MiB
[ 0.000000] OF: reserved mem: initialized node linux,cma, compatible id shared-dma-pool
[ 0.000000] CPU: All CPU(s) started in SVC mode.
[ 0.000000] percpu: Embedded 20 pages/cpu s50920 r8192 d22808 u81920
[ 0.000000] Built 1 zonelists, mobility grouping on. Total pages: 130048
[ 0.000000] Kernel command line: console=ttymxc0,115200 rdinit=/linuxrc clk_ignore_unused mfg_mmcdev=3
[ 0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[ 0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[ 0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[ 0.000000] Memory: 154912K/524288K available (11264K kernel code, 1014K rwdata, 3504K rodata, 1024K init, 6576K bss, 41696K reserved, 32
7680K cma-reserved, 0K highmem)

........

719872 bytes (720 kB, 703 KiB) copied, 0.0482173 s, 14.9 MB/s
run shell cmd: mmc=cat /tmp/mmcdev; echo 1 > /sys/block/mmcblk${mmc}boot0/force_ro
run shell cmd: mmc=cat /tmp/mmcdev; while [ ! -e /dev/mmcblk${mmc}p1 ]; do sleep 1; done
run shell cmd: mmc=cat /tmp/mmcdev; mkfs.vfat /dev/mmcblk${mmc}p1
run shell cmd: mmc=cat /tmp/mmcdev; mkdir -p /mnt/fat
run shell cmd: mmc=cat /tmp/mmcdev; mount -t vfat /dev/mmcblk${mmc}p1 /mnt/fat
WOpen:/mnt/fat
WOpen:/mnt/fat/zImage
[ 9.378323] random: crng init done
WOpen:/mnt/fat
WOpen:/mnt/fat/imx6dl-sabresd.dtb
run shell cmd: umount /mnt/fat
run shell cmd: mmc=cat /tmp/mmcdev; mkfs.ext4 /dev/mmcblk${mmc}p2
mke2fs 1.45.3 (14-Jul-2019)
run shell cmd: mmc=cat /tmp/mmcdev; mkfs.ext4 /dev/mmcblk${mmc}p3
mke2fs 1.45.3 (14-Jul-2019)
run shell cmd: mmc=cat /tmp/mmcdev; mkfs.ext4 /dev/mmcblk${mmc}p4
mke2fs 1.45.3 (14-Jul-2019)
run shell cmd: mkdir -p /mnt/rootfs
run shell cmd: mmc=cat /tmp/mmcdev; mount -t ext4 /dev/mmcblk${mmc}p2 /mnt/rootfs
[ 20.723176] EXT4-fs (mmcblk3p2): mounted filesystem with ordered data mode. Opts: (null)
run shell cmd: tar -jx -C /mnt/rootfs
WOpen:-
[ 35.050381] usb_otg_vbus: disabling
[ 35.054076] usb_h1_vbus: disabling
[ 35.057512] wm8962-supply: disabling
[ 35.071654] MPCIE_3V3: disabling
[ 35.074928] sensors-supply: disabling
[ 64.491168] imx-sdma 20ec000.sdma: Direct firmware load for imx/sdma/sdma-imx6q.bin failed with error -2
[ 64.513356] imx-sdma 20ec000.sdma: Falling back to sysfs fallback for: imx/sdma/sdma-imx6q.bin
[ 69.610247] cfg80211: failed to load regulatory.db
failure write to usb ep

[nxpfrankli]

I check your kernel git commit hash value, which not include, Latest kernel will publish at Jun 30. Please update it and try again.

commit ff01b0f10d5f5ada52140561b21387092dadf446
Author: Peter Chen peter.chen@nxp.com
Date: Fri Feb 21 21:40:57 2020 +0800

MLK-23349-2 usb: chipidea: udc: add sg list support

For low system memory system (eg, 265M), it may met OOM issue if the
request buffer is large (eg, 64KB). We ran out below OOM issue for f_fs.
Luckily, the f_fs supports sg list now, the OOM issue is fixed with
this patch.

ufb: page allocation failure: order:4, mode:0x40cc0(GFP_KERNEL|__GFP_COMP),
nodemask=(null),cpuset=/,mems_allowed=0
CPU: 2 PID: 370 Comm: ufb Not tainted 5.4.3-1.1.0+g54b3750d61fd #1
Hardware name: NXP i.MX8MNano DDR4 EVK board (DT)

[franklin-woo]

udc: add sg list support

I check your kernel git commit hash value, which not include, Latest kernel will publish at Jun 30. Please update it and try again.

commit ff01b0f10d5f5ada52140561b21387092dadf446
Author: Peter Chen peter.chen@nxp.com
Date: Fri Feb 21 21:40:57 2020 +0800

MLK-23349-2 usb: chipidea: udc: add sg list support

For low system memory system (eg, 265M), it may met OOM issue if the
request buffer is large (eg, 64KB). We ran out below OOM issue for f_fs.
Luckily, the f_fs supports sg list now, the OOM issue is fixed with
this patch.

ufb: page allocation failure: order:4, mode:0x40cc0(GFP_KERNEL|__GFP_COMP),
nodemask=(null),cpuset=/,mems_allowed=0
CPU: 2 PID: 370 Comm: ufb Not tainted 5.4.3-1.1.0+g54b3750d61fd #1
Hardware name: NXP i.MX8MNano DDR4 EVK board (DT)

Thank you for reply.

I found the patch below link.

https://www.spinics.net/lists/linux-usb/msg191815.html
https://www.spinics.net/lists/linux-usb/msg194332.html

I applied the patch in my kernel, But it's not the work. I get the same error as before. :(
Can you give me the official patch link?

udc: add sg list support

I check your kernel git commit hash value, which not include, Latest kernel will publish at Jun 30. Please update it and try again.
commit ff01b0f10d5f5ada52140561b21387092dadf446
Author: Peter Chen peter.chen@nxp.com
Date: Fri Feb 21 21:40:57 2020 +0800

MLK-23349-2 usb: chipidea: udc: add sg list support

For low system memory system (eg, 265M), it may met OOM issue if the
request buffer is large (eg, 64KB). We ran out below OOM issue for f_fs.
Luckily, the f_fs supports sg list now, the OOM issue is fixed with
this patch.

ufb: page allocation failure: order:4, mode:0x40cc0(GFP_KERNEL|__GFP_COMP),
nodemask=(null),cpuset=/,mems_allowed=0
CPU: 2 PID: 370 Comm: ufb Not tainted 5.4.3-1.1.0+g54b3750d61fd #1
Hardware name: NXP i.MX8MNano DDR4 EVK board (DT)

Thank you for reply.

I found the patch below link.

https://www.spinics.net/lists/linux-usb/msg191815.html
https://www.spinics.net/lists/linux-usb/msg194332.html

I applied the patch in my kernel, But it's not the work. I get the same error as before. :(
Can you give me the official patch link?

Hi, have you solved your issue ? im experiencing the same one under windows/ubuntu
my rootfs file size only 100MB

===============================================================
C:\Users---\Downloads\windows>uuu ./example_kernel_nand.uuu
uuu (Universal Update Utility) for nxp imx chips -- libuuu_1.3.191-0-g4fe24b9

Success 0 Failure 1

1:10 21/24 [Bulk(R):LIBUSB_ERROR_TIMEOUT ] FBK: ucp rootfs.tar t:-

and using this script ending:

burn rootfs

FBK: ucmd source /tmp/mtd.sh; flash_erase /dev/mtd${nandrootfs} 0 0
FBK: ucmd source /tmp/mtd.sh; ubiattach /dev/ubi_ctrl -m ${nandrootfs}
FBK: ucmd source /tmp/mtd.sh; ubimkvol /dev/ubi0 -Nnandrootfs -m
FBK: ucmd source /tmp/mtd.sh; mkdir -p /mnt/mtd
FBK: ucmd source /tmp/mtd.sh; mount -t ubifs ubi0:nandrootfs /mnt/mtd

FBK: acmd export EXTRACT_UNSAFE_SYMLINKS=1; tar -xv -C /mnt/mtd
FBK: ucp rootfs.tar t:-
FBK: sync
FBK: ucmd umount /mnt/mtd

FBK: done