ubiformat in barebox giving timeout - linux

I have a custom iMX 6UL board with Barebox (partially) functional. I have on board a Semper s25hs512t Flash being detected (after adding the necessary device id indrivers/mtd/spi-nor/spi-nor.c)
The problem - My board does not have ethernet or removable SD. I need to burn the boot loader/ flash on the s25hs512. I need to format the flash accordingly and copy the files on it.
my dtsi has
&qspi {
pinctrl-names = "default";
pinctrl-0 = <&pinctrl_qspi>;
status = "okay";
flash0: s25hs512t#0 {
#address-cells = <1>;
#size-cells = <1>;
compatible = "spansion,s25hs512t", "jedec,spi-nor";
spi-max-frequency = <40000000>;
spi-rx-bus-width = <4>;
spi-tx-bus-width = <4>;
reg = <0>;
spi-mode = <0>;
m25p,fast-read;
status = "okay";
partition#0 {
label = "barebox";
reg = <0x00000000 0x00100000>;
};
partition#1 {
label = "barebox-env";
reg = <0x00100000 0x00040000>;
};
partition#2 {
label = "barebox-of";
reg = <0x00140000 0x00040000>;
};
partition#3 {
label = "kernel";
reg = <0x00180000 0x00800000>;
};
partition#4 {
label = "root";
reg = <0x00980000 0x03640000>;
};
};
};
on boot barebox detects the flash
Board: Freescale i.MX6 UltraLite Caisteal Board
detected i.MX6 UltraLite revision 1.0
i.MX6 UltraLite unique ID: 241e09d4e317402a
m25p80 s25hs512t#00: s25hs512t (65536 Kbytes). <=====
imx-esdhc 2194000.mmc#2194000.of: registered as mmc1
rng_self_test: RNG software self-test passed
caam 2140000.crypto#2140000.of: Instantiated RNG4 SH0
caam 2140000.crypto#2140000.of: Instantiated RNG4 SH1
malloc space: 0x8eefcf80 -> 0x9ddf9eff (size 239 MiB)
barebox-environment chosen:environment.of: probe failed: No such file or directory
devinfo shows
`-- 21e0000.spi#21e0000.of
`-- s25hs512t#00
`-- m25p0
`-- 0x00000000-0x03ffffff ( 64 MiB): /dev/m25p0
`-- m25p0.barebox
`-- 0x00000000-0x000fffff ( 1 MiB): /dev/m25p0.barebox
`-- m25p0.barebox-env
`-- 0x00000000-0x0003ffff ( 256 KiB): /dev/m25p0.barebox-env
`-- m25p0.barebox-of
`-- 0x00000000-0x0003ffff ( 256 KiB): /dev/m25p0.barebox-of
`-- m25p0.kernel
`-- 0x00000000-0x007fffff ( 8 MiB): /dev/m25p0.kernel
`-- m25p0.root
`-- 0x00000000-0x0363ffff ( 54.3 MiB): /dev/m25p0.root
but when I run ubiformat, I am oddly getting this
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ ubiformat /dev/m25p0.barebox -y
ubiformat: m25p0.barebox (nor), size 1048576 bytes (1 MiB), 4 eraseblocks of 262144 bytes (256 KiB), min. I/O size 1 bytes
libscan: scanning eraseblock 3 -- 100 % complete
ubiformat: 1 eraseblocks are supposedly empty
ubiformat: warning!: 3 of 4 eraseblocks contain non-ubifs data
ubiformat: warning!: only 0 of 4 eraseblocks have valid erase counter
ubiformat: erase counter 0 will be used for all eraseblocks
ubiformat: note, arbitrary erase counter value may be specified using -e option
ubiformat: use erase counter 0 for all eraseblocks
ubiformat: formatting eraseblock 3 -- 100 % complete
ERROR: m25p80 s25hs512t#00: flash operation timed out
ERROR: m25p0.barebox: error -110 while writing 262144 bytes to PEB 0:0, written 0 bytes
libubigen: error!: cannot write 262144 bytes
ubiformat: error!: cannot write layout volume
ubiformat: Operation not permitted
Any way ahead from this?
PS : Update
Thanks for help from #TrentP - I am focusing only on formatting the larger partitions so that I can write the kernel and root partition. but I have not been able to mount the ubi partition. I get the following issue (Readonly filesystem)
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ erase /dev/m25p0.kernel
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ ubiattach /dev/m25p0.kernel
NOTICE: ubi0: scanning is finished
NOTICE: ubi0: empty MTD device detected
NOTICE: ubi0: registering /dev/m25p0.kernel.ubi
NOTICE: ubi0: attached mtd0 (name "m25p0.kernel", size 8 MiB) to ubi0
NOTICE: ubi0: PEB size: 262144 bytes (256 KiB), LEB size: 262016 bytes
NOTICE: ubi0: min./max. I/O unit sizes: 1/256, sub-page size 1
NOTICE: ubi0: VID header offset: 64 (aligned 64), data offset: 128
NOTICE: ubi0: good PEBs: 32, bad PEBs: 0, corrupted PEBs: 0
NOTICE: ubi0: user volume: 0, internal volumes: 1, max. volumes count: 128
NOTICE: ubi0: max/mean erase counter: 1/0, WL threshold: 65536, image sequence number: 1700878141
NOTICE: ubi0: available PEBs: 28, total reserved PEBs: 4, PEBs reserved for bad PEB handling: 0
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ ubimkvol /dev/m25p0.kernel.ubi kernel 0
NOTICE: ubi0: registering kernel as /dev/m25p0.kernel.ubi.kernel
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ mount -t ubifs /dev/m25p0.kernel.ubi.kernel /mnt/kernel/
ERROR: UBIFS error (ubi0:0): 9de5a2d5: can't format empty UBI volume: read-only mount
ERROR: ubifs ubifs0: probe failed: Read-only file system
mount: Invalid argument
If I use ubiformat I get this
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ ubiformat /dev/m25p0.kernel -y
ubiformat: m25p0.kernel (nor), size 8388608 bytes (8 MiB), 32 eraseblocks of 262144 bytes (256 KiB), min. I/O size 1 bytes
libscan: scanning eraseblock 31 -- 100 % complete
ubiformat: warning!: 32 of 32 eraseblocks contain non-ubifs data
ubiformat: warning!: only 0 of 32 eraseblocks have valid erase counter
ubiformat: erase counter 0 will be used for all eraseblocks
ubiformat: note, arbitrary erase counter value may be specified using -e option
ubiformat: use erase counter 0 for all eraseblocks
ubiformat: formatting eraseblock 31 -- 100 % complete
barebox#Freescale i.MX6 UltraLite Caisteal Board:/ ubiattach /dev/m25p0.kernel
NOTICE: ubi0: scanning is finished
ERROR: ubi0 error: ubi_read_volume_table: the layout volume was not found
ERROR: ubi0 error: ubi_attach_mtd_dev: failed to attach mtd0, error -22
failed to attach: Invalid argument
devinfo
Parent: m25p0.kernel
Parameters:
available_pebs: 0 (type: uint32)
bad_peb_count: 0 (type: uint32)
good_peb_count: 32 (type: uint32)
leb_size: 262016 (type: uint32)
max_erase_counter: 2 (type: uint32)
mean_erase_counter: 0 (type: uint32)
min_io_size: 1 (type: uint32)
peb_size: 262144 (type: uint32)
reserved_pebs: 32 (type: uint32) <=== why all PEBs are reserved?
sub_page_size: 1 (type: uint32)
vid_header_offset: 64 (type: uint32)
Any suggestions on what I am doing wrong. I know its something ridiculously simple. just unknown to me

You aren't supposed to use ubiformat on the barebox partition. It's too small. That's why it fails.
UBI is a Linux layer for putting UBI filesystems into NAND or NOR flash. The iMX6UL CPU boot ROM does not understand UBI. It can't boot something in a UBI formatted partition. It's for the root filesystem in the root partition.
Read section 8 of the iMX6UL reference manual, especially §8.6 about QuadSPI booting. This will tell you what you must put into flash to make it bootable.
Also look at the barebox_update command, which can be used to flash the bootloader from Barebox. The board needs to support it and I don't know about your board. The code is in various imx6_bbu_* functions. I'm not sure if qspi is supported, as I only see eMMC/SD,eMMC boot, NAND, and I2C/SPI. The qspi interface isn't the same as a serial EEPROM on one of the eCSPI controllers (again, see RM §8!). But perhaps it would work with an appropriate header already on the image.

Related

u-boot gives Error 22 for ubi partition, but mounts ok in linux

I have a buildroot system, which mounts ubi ok in linux, but in u-boot I get error 22
When starting in linux this is in dmesg:
ubi0: scanning is finished
ubi0: attached mtd2 (name "rootfs", size 32 MiB)
ubi0: PEB size: 131072 bytes (128 KiB), LEB size: 126976 bytes
ubi0: min./max. I/O unit sizes: 2048/2048, sub-page size 2048
ubi0: VID header offset: 2048 (aligned 2048), data offset: 4096
ubi0: good PEBs: 256, bad PEBs: 0, corrupted PEBs: 0
ubi0: user volume: 1, internal volumes: 1, max. volumes count: 128
ubi0: max/mean erase counter: 2/0, WL threshold: 4096, image sequence number: 894512245
ubi0: available PEBs: 0, total reserved PEBs: 256, PEBs reserved for bad PEB handling: 40
ubi0: background thread "ubi_bgt0d" started, PID 1103
--
UBIFS (ubi0:0): UBIFS: mounted UBI device 0, volume 0, name "rootfs", R/O mode
UBIFS (ubi0:0): LEB size: 126976 bytes (124 KiB), min./max. I/O unit sizes: 2048 bytes/2048 bytes
UBIFS (ubi0:0): FS size: 25649152 bytes (24 MiB, 202 LEBs), journal size 4444160 bytes (4 MiB, 35 LEBs)
UBIFS (ubi0:0): reserved for root: 0 bytes (0 KiB)
UBIFS (ubi0:0): media format: w4/r0 (latest is w4/r0), UUID 29B5D4CF-8B0B-465A-8D03-F3A464E6250E, small LPT model
UBIFS (ubi0:0): full atime support is enabled.
VFS: Mounted root (ubifs filesystem) readonly on device 0:13.
in u-boot mtd returns:
device nand0 <nand0>, # parts = 4
#: name size offset mask_flags
0: u-boot 0x00200000 0x00000000 0
1: kernel 0x01e00000 0x00200000 0
2: rootfs 0x02000000 0x02000000 0
3: user 0x0c000000 0x04000000 0
active partition: nand0,0 - (u-boot) 0x00200000 # 0x00000000
defaults:
mtdids : nand0=nand0
mtdparts: mtdparts=nand0:0x200000#0x0(u-boot),0x1e00000#0x200000(kernel),0x2000000#0x2000000(rootfs),-(user)
but when it try to attach:
=> ubi part rootfs
ubi0: attaching mtd1
UBI init error 22
It's on an embedded system which uses older versions U-Boot 2016.11 and Linux/arm 4.4.289 Kernel
I suppose some parameter is wrong somewhere, can somebody give me some advise where to look?

Linux Kernel DTSI File Cannot Compile from Yocto for BeagleBone

I am using the Texas Instruments official Yocto SDK to build a complete BSP for the Beaglebone X-15 ( TI AM5728 Processor ).
The entire SDK builds great for the MACHINE=am57xx-evm type from the SDK. Later SDKs include the MACHINE=beagle-x15, but I need this older version with Linux kernel 4.4.
The Linux kernel 4.4 in this SDK does have beagle-x15 device tree fragments included, but the machine configuration for the beagle x15 was not present, so I included the 1 conf file for the new machine from a later SDK.
The problem is that the Device Tree fails to compile - there is a syntax issue as shown in this traceback:
| Error: /home/user/tisdk/build/arago-tmp-external-linaro-toolchain/work-shared/beagle-x15/kernel-source/arch/arm/boot/dts/beagle-x15-cmem.dtsi:1.1-2 syntax error
| FATAL ERROR: Unable to parse input tree
| scripts/Makefile.lib:293: recipe for target 'arch/arm/boot/dts/am57xx-beagle-x15-revc.dtb' failed
| make[3]: *** [arch/arm/boot/dts/am57xx-beagle-x15-revc.dtb] Error 1
| arch/arm/Makefile:333: recipe for target 'am57xx-beagle-x15-revc.dtb' failed
| make[2]: *** [am57xx-beagle-x15-revc.dtb] Error 2
| Makefile:150: recipe for target 'sub-make' failed
| make[1]: *** [sub-make] Error 2
| Makefile:24: recipe for target '__sub-make' failed
| make: *** [__sub-make] Error 2
| WARNING: /home/user/tisdk/build/arago-tmp-external-linaro-toolchain/work/beagle_x15-linux-gnueabi/linux-ti-staging/4.4.41+gitAUTOINC+f9f6f0db2d-r7a.arago5.tisdk60/temp/run.do_compile.121513:1 exit 1 from 'exit 1'
| ERROR: oe_runmake failed
| ERROR: Function failed: do_compile (log file is located at /home/user/tisdk/build/arago-tmp-external-linaro-toolchain/work/beagle_x15-linux-gnueabi/linux-ti-staging/4.4.41+gitAUTOINC+f9f6f0db2d-r7a.arago5.tisdk60/temp/log.do_compile.121513)
Here is the entire DTSI file that fails to compile:
/ {
reserved-memory {
#address-cells = <2>;
#size-cells = <2>;
ranges;
cmem_block_mem_0: cmem_block_mem#a0000000 {
reg = <0x0 0xa0000000 0x0 0x0c000000>;
no-map;
status = "okay";
};
cmem_block_mem_1_ocmc3: cmem_block_mem#40500000 {
reg = <0x0 0x40500000 0x0 0x100000>;
no-map;
status = "okay";
};
};
cmem {
compatible = "ti,cmem";
#address-cells = <1>;
#size-cells = <0>;
#pool-size-cells = <2>;
status = "okay";
cmem_block_0: cmem_block#0 {
reg = <0>;
memory-region = <&cmem_block_mem_0>;
cmem-buf-pools = <1 0x0 0x0c000000>;
};
cmem_block_1: cmem_block#1 {
reg = <1>;
memory-region = <&cmem_block_mem_1_ocmc3>;
};
};
};
Here is the beagle-x15.conf file:
##TYPE: Machine
##NAME: BeagleBoard X15
##DESCRIPTION: Machine configuration for the BeagleBoard X15
require conf/machine/include/dra7xx.inc
KERNEL_DEVICETREE = "am57xx-beagle-x15.dtb am57xx-beagle-x15-revb1.dtb am57xx-beagle-x15-revc.dtb"
MACHINE_GUI_CLASS = "bigscreen"
SERIAL_CONSOLE = "115200 ttyS2"
UBOOT_MACHINE = "am57xx_evm_config"
WKS_FILE = "sdimage-bootpart.wks"
IMAGE_BOOT_FILES = "MLO u-boot.img"
IMAGE_FSTYPES += "tar.xz wic.xz"
do_image_wic[depends] += "mtools-native:do_populate_sysroot dosfstools-native:do_populate_sysroot"
# UBI information. Note that this is board and kernel specific. Changes
# in your kernel port may require changes in these variables. For more
# details about this board please see
# http://processors.wiki.ti.com/index.php/UBIFS_Support
# do ubiattach /dev/ubi_ctrl -m 7 -O 2048
# From dmesg:
# UBI: smallest flash I/O unit: 2048
# UBI: logical eraseblock size: 126976 bytes
# from ubiattach stdout:
# UBI device number 0, total 1988 LEBs
MKUBIFS_ARGS = "-F -m 2048 -e 126976 -c 8192"
# do ubiattach /dev/ubi_ctrl -m 7 -O 2048
# from dmesg:
# UBI: smallest flash I/O unit: 2048
# UBI: physical eraseblock size: 131072 bytes (128 KiB)
# UBI: sub-page size: 512
# UBI: VID header offset: 2048 (aligned 2048)
UBINIZE_ARGS = "-m 2048 -p 128KiB -s 512 -O 2048"
How can I get this DTSI file to compile? Thanks.
UPDATE: It turns out the Ubuntu DTC compiler also fails ( Version 1.4 ):
dtc -O dtb -o /home/user/Desktop/test.dtb /home/user/tisdk/build/arago-tmp-external-linaro-toolchain/work-shared/beagle-x15/kernel-source/arch/arm/boot/dts/beagle-x15-cmem.dtsi
Error: /home/user/tisdk/build/arago-tmp-external-linaro-toolchain/work-shared/beagle-x15/kernel-source/arch/arm/boot/dts/beagle-x15-cmem.dtsi:1.1-2 syntax error
FATAL ERROR: Unable to parse input tree
dtc compiler by default treats the Device Tree version as 0 if no version is specified. Syntax for version 0 is different from version 1. So you need to add,
/dts-v1/;
as your fist line of device tree file.
Apart from this, usually you need to compile .dts file, not the .dtsi (which is include) directly. So you need to define according device tree file with .dts extension including .dtsi files.

Linux SLUB: Unable to allocate memory on node

We are getting very frequently below message in /var/log/messages
kernel: SLUB: Unable to allocate memory on node -1 (gfp=0x8020)
In some cases followed by an allocation table
kernel: cache: sigqueue(12019:454c4ebd186d964699132181ad7367c669700f7d8991c47d4bc053ed101675bc), object size: 160, buffer size: 160, default order: 0, min order: 0
kernel: node 0: slabs: 57, objs: 23313, free: 0
kernel: node 1: slabs: 35, objs: 14315, free: 0
Ok, free is 0, but how may this be tuned?
Following is set information
OS - Centos7.3
Kernel - 3.10.0-327.36.3.el7.x86_64
Docker - 1.12.6
Kubernetes - 1.5.5
We have private cloud powered by kurbernetes, having 10 nodes; it was working fine till last month and now we are getting these alerts very frequently on every nodes, pods/container also increased in last few days.
We have enough memory and cpu available on each node.
Any fine tuning for these alert will be very helpful.
Additional information:
sysctl.conf options
net.ipv4.tcp_timestamps = 0
net.ipv4.tcp_max_syn_backlog = 4096
net.core.somaxconn = 1024
net.ipv4.tcp_syncookies = 1
net.core.rmem_max = 16777216
net.core.wmem_max = 16777216
net.core.rmem_default = 65535
net.core.wmem_default = 65535
net.ipv4.tcp_rmem = 4096 87380 16777216
net.ipv4.tcp_wmem = 4096 65536 16777216
net.ipv4.ip_local_port_range = 1024 65535
vm.max_map_count = 262144
vm.swappiness=10
vm.vfs_cache_pressure=100
Please look at this: https://pingcap.com/blog/try-to-fix-two-linux-kernel-bugs-while-testing-tidb-operator-in-k8s/. It's a kernel bug.
problems seems to be with kernel, first a fall check whether swap memory is properly allocated or not by free -m and mkswap -c, if swap is not properly allocated, do it. if swap is fine, then you might need to update the kernel.

Halide AOT for OpenCL works fine as static library but not as shared object

I try to compile the code below both to static library and to object file:
Halide::Func f("f");
Halide::Var x("x");
f(x) = x;
f.gpu_tile(x, 4);
f.bound(x, 0, 16);
Halide::Target target = Halide::get_target_from_environment();
target.set_feature(Halide::Target::OpenCL);
target.set_feature(Halide::Target::Debug);
// f.compile_to_static_library("mylib", {}, "f", target);
// f.compile_to_file("mylib", {}, "f", target);
In case of static linking all works fine and output result is correct:
Halide::Buffer<int> output(16);
f(output.raw_buffer());
output.copy_to_host();
std::cout << output(10) << std::endl;
But when I try link object file into shared object,
gcc -shared -pthread mylib.o -o mylib.so
And open it from code (Ubuntu 16.04),
void* handle = dlopen("mylib.so", RTLD_NOW);
int (*func)(halide_buffer_t*);
*(void**)(&func) = dlsym(handle, "f");
func(output.raw_buffer());
I receive CL_INVALID_MEM_OBJECT error. Here is the debugging log:
CL: halide_opencl_init_kernels (user_context: 0x0, state_ptr: 0x7f1266b5a4e0, program: 0x7f1266957480, size: 1577
load_libopencl (user_context: 0x0)
Loaded OpenCL runtime library: libOpenCL.so
create_opencl_context (user_context: 0x0)
Got platform 'Intel(R) OpenCL', about to create context (t=6249430)
Multiple CL devices detected. Selecting the one with the most cores.
Device 0 has 20 cores
Device 1 has 4 cores
Selected device 0
device name: Intel(R) HD Graphics
device vendor: Intel(R) Corporation
device profile: FULL_PROFILE
global mem size: 1630 MB
max mem alloc size: 815 MB
local mem size: 65536
max compute units: 20
max workgroup size: 256
max work item dimensions: 3
max work item sizes: 256x256x256x0
clCreateContext -> 0x1899af0
clCreateCommandQueue 0x1a26a80
clCreateProgramWithSource -> 0x1a26ab0
clBuildProgram 0x1a26ab0 -D MAX_CONSTANT_BUFFER_SIZE=854799155 -D MAX_CONSTANT_ARGS=8
Time: 1.015832e+02 ms
CL: halide_opencl_run (user_context: 0x0, entry: kernel_f_s0_x___deprecated_block_id_x___block_id_x, blocks: 4x1x1, threads: 4x1x1, shmem: 0
clCreateKernel kernel_f_s0_x___deprecated_block_id_x___block_id_x -> Time: 1.361700e-02 ms
clSetKernelArg 0 4 [0x2e00010000000000 ...] 0
clSetKernelArg 1 8 [0x2149040 ...] 1
Mapped dev handle is: 0x2149040
Error: CL: clSetKernelArg failed: CL_INVALID_MEM_OBJECT
Aborted (core dumped)
Thank you very much for help! Commit state c7375fa. I'm pleasure provide extra information if it will be necessary.
Solution: In this case we have runtime duplication. Load shared object with flag RTLD_DEEPBIND.
void* handle = dlopen("mylib.so", RTLD_NOW | RTLD_DEEPBIND);
RTLD_DEEPBIND (since glibc 2.3.4)
Place the lookup scope of the symbols in this library ahead of the global scope. This means that a self-contained library will use its own symbols in preference to global symbols with the same name contained in libraries that have already been loaded. This flag is not specified in POSIX.1-2001.
https://linux.die.net/man/3/dlopen

Can the logical erase block size of an MTD device be increased?

The minimum erase block size for jffs2 (mtd-utils version 1.5.0, mkfs.jffs2 revision 1.60) seems to be 8KiB:
Erase size 0x1000 too small. Increasing to 8KiB minimum
However I am running Linux 3.10 with an at25df321a,
m25p80 spi32766.0: at25df321a (4096 Kbytes),
and the erase block size is only 4KiB:
mtd5
Name: spi32766.0
Type: nor
Eraseblock size: 4096 bytes, 4.0 KiB
Amount of eraseblocks: 1024 (4194304 bytes, 4.0 MiB)
Minimum input/output unit size: 1 byte
Sub-page size: 1 byte
Character device major/minor: 90:10
Bad blocks are allowed: false
Device is writable: true
Is there a way to make the mtd system treat multiple erase blocks as one? Maybe some ioctl or module parameter?
If I flash a jffs2 image with larger erase block size, I get lots of kernel error messages, missing files and sometimes panic.
workaround
I found that flasherase --jffs2 results in a working filesystem inspite of the 4KiB erase block size. So I hacked the mkfs.jfss2.c file and the resulting image seems to work fine. I'll give it some testing.
diff -rupN orig/mkfs.jffs2.c new/mkfs.jffs2.c
--- orig/mkfs.jffs2.c 2014-10-20 15:43:31.751696500 +0200
+++ new/mkfs.jffs2.c 2014-10-20 15:43:12.623431400 +0200
## -1659,11 +1659,11 ## int main(int argc, char **argv)
}
erase_block_size *= units;
- /* If it's less than 8KiB, they're not allowed */
- if (erase_block_size < 0x2000) {
- fprintf(stderr, "Erase size 0x%x too small. Increasing to 8KiB minimum\n",
+ /* If it's less than 4KiB, they're not allowed */
+ if (erase_block_size < 0x1000) {
+ fprintf(stderr, "Erase size 0x%x too small. Increasing to 4KiB minimum\n",
erase_block_size);
- erase_block_size = 0x2000;
+ erase_block_size = 0x1000;
}
break;
}
http://lists.infradead.org/pipermail/linux-mtd/2010-September/031876.html
JFFS2 should be able to fit at least one node to eraseblock. The
maximum node size is 4KiB+few bytes. This is why the minimum
eraseblocks size is 8KiB.
But in practice, even 8KiB is bad because you and up with wasting a
lot of space at the end of eraseblocks.
You should join several erasblock into one virtual eraseblock of 64 or
128 KiB and use it - this will be more optimal.
Some drivers have already implemented this. I know about
MTD_SPI_NOR_USE_4K_SECTORS
Linux configuration option. It have to be set to "n" to enable large erase sectors of size 0x00010000.

Resources