On Sep 10, 2021 / 15:56, Jaegeuk Kim wrote:
> This patch fixes the below stale magic info.
>
> $ mkfs.btrfs -f test.img
> btrfs-progs v5.10.1
> See http://btrfs.wiki.kernel.org for more information.
>
> Label: (null)
> UUID: 941d2db7-3ece-4090-8b22-c4ea548b5dae
> Node size: 16384
> Sector size: 4096
> Filesystem size: 1.00GiB
> Block group profiles:
> Data: single 8.00MiB
> Metadata: DUP 51.19MiB
> System: DUP 8.00MiB
> SSD detected: no
> Incompat features: extref, skinny-metadata
> Runtime features:
> Checksum: crc32c
> Number of devices: 1
> Devices:
> ID SIZE PATH
> 1 1.00GiB test.img
>
> $ hexdump -s 0x10000 -n 128 -C test.img
> 00010000 81 29 94 0a 00 00 00 00 00 00 00 00 00 00 00 00 |.)..............|
> 00010010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
> 00010020 94 1d 2d b7 3e ce 40 90 8b 22 c4 ea 54 8b 5d ae |..-.>.@.."..T.].|
> 00010030 00 00 01 00 00 00 00 00 01 00 00 00 00 00 00 00 |................|
> 00010040 5f 42 48 52 66 53 5f 4d 05 00 00 00 00 00 00 00 |_BHRfS_M........|
> 00010050 00 40 d2 01 00 00 00 00 00 40 50 01 00 00 00 00 |.@.......@P.....|
> 00010060 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
> 00010070 00 00 00 40 00 00 00 00 00 00 02 00 00 00 00 00 |...@............|
> 00010080
>
> $ mkfs.f2fs -t 0 -f test.img
>
> F2FS-tools: mkfs.f2fs Ver: 1.14.0 (2020-12-28)
>
> Info: Disable heap-based policy
> Info: Debug level = 0
> Info: Trim is disabled
> Info: Segments per section = 1
> Info: Sections per zone = 1
> Info: sector size = 512
> Info: total sectors = 2097152 (1024 MB)
> Info: zone aligned segment0 blkaddr: 512
> Info: format version with
> "Linux version 5.10.46-4rodete1-amd64 ([email protected]) (gcc-10
> (Debian 10.2.1-6+build2) 10.2.1 20210110, GNU ld (GNU Binutils for Debian)
> 2.35.2) #1 SMP Debian 5.10.46-4rodete1 (2021-08-20)"
> Info: Overprovision ratio = 6.360%
> Info: Overprovision segments = 68 (GC reserved = 39)
> Info: format successful
>
> $hexdump -s 0x10000 -n 128 -C test.img
> 00010000 c2 8a c8 26 00 00 00 00 00 00 00 00 00 00 00 00 |...&............|
> 00010010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
> 00010020 92 ab 3f c6 b7 82 49 5e 93 23 e8 c9 e9 45 7d ac |..?...I^.#...E}.|
> 00010030 00 00 01 00 00 00 00 00 01 00 00 00 00 00 00 00 |................|
> 00010040 5f 42 48 52 66 53 5f 4d 05 00 00 00 00 00 00 00 |_BHRfS_M........|
> 00010050 00 40 d2 01 00 00 00 00 00 40 50 01 00 00 00 00 |.@.......@P.....|
> 00010060 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
> 00010070 00 00 00 40 00 00 00 00 00 00 02 00 00 00 00 00 |...@............|
> 00010080
>
> --- After this patch ---
> $ mkfs.f2fs -t 0 -f test.img
>
> F2FS-tools: mkfs.f2fs Ver: 1.14.0 (2020-12-28)
>
> Info: Disable heap-based policy
> Info: Debug level = 0
> Info: Trim is disabled
> test.img appears to contain an existing filesystem (btrfs).
> Info: Segments per section = 1
> Info: Sections per zone = 1
> Info: sector size = 512
> Info: total sectors = 2097152 (1024 MB)
> Info: zone aligned segment0 blkaddr: 512
> Info: format version with
> "Linux version 5.10.46-4rodete1-amd64 ([email protected]) (gcc-10
> (Debian 10.2.1-6+build2) 10.2.1 20210110, GNU ld (GNU Binutils for Debian)
> 2.35.2) #1 SMP Debian 5.10.46-4rodete1 (2021-08-20)"
> Info: Overprovision ratio = 6.360%
> Info: Overprovision segments = 68 (GC reserved = 39)
> Info: format successful
>
> $ hexdump -s 0x10000 -n 128 -C test.img
> 00010000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
> *
> 00010080
>
> Signed-off-by: Jaegeuk Kim <[email protected]>
> ---
> fsck/main.c | 2 +-
> include/f2fs_fs.h | 1 +
> lib/libf2fs.c | 6 ++++++
> mkfs/f2fs_format_main.c | 26 ++++++++++++++++++++++----
> 4 files changed, 30 insertions(+), 5 deletions(-)
>
> diff --git a/fsck/main.c b/fsck/main.c
> index eda399cf0679..e4cfdf443867 100644
> --- a/fsck/main.c
> +++ b/fsck/main.c
> @@ -1120,7 +1120,7 @@ int main(int argc, char **argv)
> }
>
> /* Get device */
> - if (f2fs_get_device_info() < 0) {
> + if (f2fs_get_device_info() < 0 || f2fs_get_f2fs_info() < 0) {
> ret = -1;
> if (c.func == FSCK)
> ret = FSCK_OPERATIONAL_ERROR;
> diff --git a/include/f2fs_fs.h b/include/f2fs_fs.h
> index 2a9c1169456c..69260a6fccd9 100644
> --- a/include/f2fs_fs.h
> +++ b/include/f2fs_fs.h
> @@ -1341,6 +1341,7 @@ extern int f2fs_devs_are_umounted(void);
> extern int f2fs_dev_is_writable(void);
> extern int f2fs_dev_is_umounted(char *);
> extern int f2fs_get_device_info(void);
> +extern int f2fs_get_f2fs_info(void);
> extern unsigned int calc_extra_isize(void);
> extern int get_device_info(int);
> extern int f2fs_init_sparse_file(void);
> diff --git a/lib/libf2fs.c b/lib/libf2fs.c
> index b27785dec757..420dfdae56da 100644
> --- a/lib/libf2fs.c
> +++ b/lib/libf2fs.c
> @@ -1183,6 +1183,12 @@ int f2fs_get_device_info(void)
> for (i = 0; i < c.ndevs; i++)
> if (get_device_info(i))
> return -1;
> + return 0;
> +}
> +
> +int f2fs_get_f2fs_info(void)
> +{
> + int i;
>
> if (c.wanted_total_sectors < c.total_sectors) {
> MSG(0, "Info: total device sectors = %"PRIu64" (in %u bytes)\n",
> diff --git a/mkfs/f2fs_format_main.c b/mkfs/f2fs_format_main.c
> index 031244d5d67a..de93f1b5b778 100644
> --- a/mkfs/f2fs_format_main.c
> +++ b/mkfs/f2fs_format_main.c
> @@ -396,18 +396,36 @@ int main(int argc, char *argv[])
>
> c.func = MKFS;
>
> - if (!force_overwrite && f2fs_check_overwrite()) {
> - MSG(0, "\tUse the -f option to force overwrite.\n");
> + if (f2fs_get_device_info() < 0)
> return -1;
Jaegeuk, thanks for the patch. Today, I had a chance to try out the dev branch
with this patch, and observed mkfs.f2fs for block devices failed with the
messages as follows:
$ sudo mkfs.f2fs -f /dev/sdc
F2FS-tools: mkfs.f2fs Ver: 1.14.0 (2021-09-10)
Info: Disable heap-based policy
Info: Debug level = 0
Info: Trim is enabled
Info: [/dev/sdc] Disk Model: SanDisk SDSSDH32
/dev/sdc appears to contain an existing filesystem (f2fs).
Error: In use by the system!
I think f2fs_get_device_info() call above opens the block device with O_EXCL.
And the following f2fs_devs_are_umounted() opens the device again with O_EXCL.
The second open results in E_BUSY and the error message. I tried to move
the f2fs_devs_are_umounted() call before the f2fs_get_device_info() call, then
the error disappeared. So, I suggest to move the f2fs_devs_are_umounted() call.
> +
> + if (f2fs_check_overwrite()) {
> + char *zero_buf = NULL;
> + int i;
> +
> + if (!force_overwrite) {
> + MSG(0, "\tUse the -f option to force overwrite.\n");
> + goto err_format;
> + }
> + zero_buf = calloc(F2FS_BLKSIZE, 1);
> + if (!zero_buf) {
> + MSG(0, "\tFaile to allocate zero buffer.\n");
I guess you meant "Failed" in place of "Faile".
> + goto err_format;
> + }
> + /* wipe out other FS magics mostly first 4MB space */
> + for (i = 0; i < 1024; i++)
> + dev_fill_block(zero_buf, i);
> + free(zero_buf);
> + f2fs_fsync_device();
> }
>
> if (f2fs_devs_are_umounted() < 0) {
> if (errno != EBUSY)
> MSG(0, "\tError: Not available on mounted device!\n");
> - return -1;
> + goto err_format;
> }
>
> - if (f2fs_get_device_info() < 0)
> + if (f2fs_get_f2fs_info() < 0)
> goto err_format;
>
> /*
> --
> 2.33.0.309.g3052b89438-goog
>
>
>
> _______________________________________________
> Linux-f2fs-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
--
Best Regards,
Shin'ichiro Kawasaki
_______________________________________________
Linux-f2fs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
