ya-mouse / fatresize Goto Github PK

View Code? Open in Web Editor NEW

13.0 13.0 21.0 233 KB

Resize FAT partitions using libparted

License: GNU General Public License v3.0

Makefile 30.95% M4 1.10% Shell 46.73% C 19.19% Roff 2.02%

fatresize's People

Contributors

Stargazers

Watchers

Forkers

xambroz pchauncey sriemer thelittlebug ptrnine mikebdp2 artoria2e5 jjmcdn mentorembedded persianyagami90xs fgroenendijk southoz hughr rhn robang74 seanmollet jwatt camio whitty paaguti

fatresize's Issues

fatresize 1.1.0 (20200405) wrong size when maximizing partition

Hello,
this is my first bug report so please correct me if I am doing something wrong. I am using Arch Linux with fatresize 1.1.0 (20200405) and tried to maximize a FAT filesystem inside a 4GiB partition on a 8GiB virtual disk created with qemu-nbd. I've used the following command:
sudo fatresize -ps max /dev/nbd0p1

However this fails with the following error:
fatresize 1.1.0 (20200405)
.part(start=2048, end=8390655, length=8388608)
Error: The location 8591MB is outside of the device /dev/nbd0.

I really can't explain why this happens, because it event prints out the correct start and end of the partition, but then seems like it is somehow trying to resize the partition to 8591MB.

Please would you have a look at this :)

Cannot resize FAT partitions on SD cards (like e.g. /dev/mmcblk0p1)

Hi,

I'm the current SUSE parted maintainer and it has been reported to me that calling fatresize with /dev/mmcblk0p1 fails (openSUSE bug: https://bugzilla.suse.com/show_bug.cgi?id=959181).

I can reproduce this with latest master commit 46cf019.

# ./fatresize -s 5G /dev/mmcblk0p1
fatresize 1.0.3 ()
Error: Could not stat device /dev/mmcblk0p - No such file or directory.

Parted supports those devices properly. Can you please develop a fix? TIA

FAT12 support

I need to resize a floppy image from 1440KB (1474560 bytes) "High Density" aka standard floppy to 2880KB (2949120 bytes) "Extended Density" floppy, and sadly discovered there are no opensource Linux tools for that! Manual hexedit of 0x13/0x14 bytes from 0x0B40 to 0x1680 and 0x18 byte from 0x12 to 0x24 - also didn't work for me (but I just learned about FAT structure and may be missing something). + #15

Please add FAT12 resizing support - it may be really important in some situations.

Wrong partition is selected from command parameter

When trying to resize a partition other than 1 on a real disk the first partition is selected instead

# fatresize /dev/mmcblk0p2 -vvv -s 800M
fatresize 1.1.0 (20201114)
ped_device_get(/dev/mmcblk0)
ped_device_open()
ped_disk_new()
ped_disk_get_partition(1)
Error: Partition /dev/mmcblk0p1 is being used.  You must unmount it before you modify it with Parted.

Workaround is to specify -n 2 in this case.

Found this as a regression to 1.0.2-11 when updating from an older yocto project.

I think this relates to #24

Changelog out of date

It looks like the changelog in the repo is out of date, as of 321973b (see here) – should this be altered to favor GitHub's releases system, maybe, or else updated with the recent information? (1.0.3 and 1.0.4 have been released, it seems, whereas the file has the most recent entry at "planned 1.0.3".)

Can't shrink fat16 partition

I'm trying to shrink my EFI FAT16 partition by 1 MB to make some space for the bcache superblock.

$ sudo parted /dev/vda
GNU Parted 3.1
Using /dev/vda
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) unit s
(parted) print
Model: Virtio Block Device (virtblk)
Disk /dev/vda: 52428800s
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags:

Number  Start   End        Size       File system  Name                  Flags
128     2048s   22527s     20480s     fat16        EFI System Partition  boot
 1      22528s  52428766s  52406239s  xfs          Linux

But fatresize fails to detect the correct boundaries and size of the partition

$ sudo ./fatresize /dev/vda -n 128 -i
fatresize 1.0.3 (20230501)
part(start=2048, end=22527, length=20480)
FAT: fat16
Cur size: 10485760
Min size: 10518528
Max size: 26843545600

And the resize fails:

$ sudo ./fatresize /dev/vda -n 128 -s 9Mi
fatresize 1.0.3 (20230501)
part(start=2048, end=22527, length=20480)
Error: Unable to satisfy all constraints on the partition.

I'm running the latest version of fatresize from this repo, compiled from source code. The OS is Amazon Linux 2 aarch64, running locally in a KVM VM.

$ ./fatresize --version
./fatresize: unrecognized option '--version'
fatresize 1.0.3 (20230501)
Usage: fatresize [options] device (e.g. /dev/hda1, /dev/sda2)
    Resize an FAT16/FAT32 volume non-destructively:

    -s, --size SIZE      Resize volume to SIZE[k|M|G|ki|Mi|Gi] bytes or "max"
    -i, --info           Show volume information
    -f, --force-yes      Do not ask questions
    -n, --partition NUM  Specify partition number
    -p, --progress       Show progress
    -q, --quiet          Be quiet
    -v, --verbose        Verbose
    -h, --help           Display this help

Please report bugs to [email protected]

Shrinking doesn't move last sector

After letting it seemingly successfully shrink a FAT32 partition on an microSD card,

fsck.vfat afterwards tells me it couldn't seek to the last sector and aborts.
Windows Checkdisk doesn't correct it but also doesn't show any detected inconsistencies or errors.
Should fatresize rewrite the last sector metadata and is it a bug/regression, or have I misunderstood something in the documentation?

segfault when trying to resize an image of a fat partition

kodi@arch-msi ~/Documents » sudo fatresize -i ./blue3ds.img
fatresize 1.1.0 (20200405)
part(start=0, end=124702719, length=124702720)
FAT: fat32
Cur size: 63847792640
Min size: 43381676032
Max size: 63847792640
kodi@arch-msi ~/Documents » sudo fatresize --size 45381676032 ./blue3ds.img
fatresize 1.1.0 (20200405)
part(start=0, end=124702719, length=124702720)
zsh: segmentation fault  sudo fatresize --size 45381676032 ./blue3ds.img

Examples broken since introducing -n

The -n option is required now, because not using it means that the program will crash (it expects getting a disk image, not a partition). This makes all examples broken, because they don't use -n. The man page is also outdated.

I "fixed" a bit of it enough to work, but it's not a proper solution: rhn@86f3ac1

Segfault when specifying maximum size (FAT_ASSERT deactivated, required action ignored)

If I get the maximum size from fatresize -i and want to set it, then fatresize segfaults in try_snap(&end, end_range,.

# fatresize -i /dev/sdb1
FAT: fat32
Size: 31103909888
Min size: 536870912
Max size: 31104958464

# fatresize -s 31104958464 /dev/sdb1
fatresize 1.0.3 ()
Segmentation fault

dmesg + objdump -S /usr/sbin/fatresize | less points to try_snap(&end, end_range,.

Regression: Cannot specify maximum size

In version 0.1 it was possible to resize the FAT partition to a maximum size e.g. by:

fatresize /dev/sdb1

Version 1.0.3 always requires a size to be set and complains if it is outside of the device.

fatresize -i /dev/sdb1

already knows the maximum size.

Would be nice to have the following:

fatresize -s max /dev/sdb1

Correct VFAT partition is not recognised

I have an ebook reader unbricking image which I transfer to a microSD card.
Using GPARTED I can grow the DOS FAT32 partition to take up all the free space.
Once grown I'm not able to read the partition's information

 paag@840g5  ~/tmp/kobo  sudo fdisk -l /dev/sda
Disk /dev/sda: 7,4 GiB, 7948206080 bytes, 15523840 sectors
Disk model: SD/MMC          
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: 0x66e8b073

Device     Boot   Start      End  Sectors  Size Id Type
/dev/sda1         19456   543744   524289  256M 83 Linux
/dev/sda2        543745  1068033   524289  256M 83 Linux
/dev/sda3       1068034 15523839 14455806  6,9G  b W95 FAT32

paag@840g5  ~/tmp/kobo  sudo fatresize -i /dev/sda3
fatresize 1.1.0 (20201114)
.Error: /dev/sda3 is not valid FAT16/FAT32 partition.

sudo fsck.vfat /dev/sda3             
fsck.fat 4.2 (2021-01-31)
There are differences between boot sector and its backup.
This is mostly harmless. Differences: (offset:original/backup)
  65:00/01
1) Copy original to backup
2) Copy backup to original
3) No action
[123?q]? 1

*** Filesystem was changed ***
The changes have not yet been written, you can still choose to leave the
filesystem unmodified:
1) Write changes
2) Leave filesystem unchanged
[12?q]? 1
/dev/sda3: 84 files, 1444/1803444 clusters

Misleading error message when permissions to device are lacking

When executing fatresize as a user that lacks permissions to the specified device, it fails with the following message:

You must specify exactly one existing device.

Since the user specified a device, this message is misleading. I ran into this issue and I suspect the two users participating in this reddit thread did as well.

The root cause seems to be the ped_device_get call in the probe_device function which errors out due to lack of permissions, but the error message is swallowed instead of reported back to the user.

I'll create a PR shortly that doesn't address the root cause, but at least provides an error message that hints to the user that their lack of permissions may be the problem.

Bug: Assertion (disk != NULL) at ../../libparted/disk.c:1620 in function ped_disk_get_partition_by_sector() failed

Got this 'bug' (according to the program itself) while trying to shrink a DD image of a Raspberry Pi boot partition.

/usr/sbin/fatresize -v -s 200M -p boot_partition.fs
fatresize 1.1.0 (20201114)
.part(start=0, end=524287, length=524288)
Backtrace has 5 calls on stack:
  5: /lib/aarch64-linux-gnu/libparted.so.2(ped_assert+0x54) [0xffffb7156b34]
  4: /lib/aarch64-linux-gnu/libparted.so.2(ped_disk_max_partition_length+0) [0xffffb715b870]
  3: /usr/sbin/fatresize(+0x2494) [0xaaaadff02494]
  2: /lib/aarch64-linux-gnu/libc.so.6(__libc_start_main+0xe8) [0xffffb6ff3e18]
  1: /usr/sbin/fatresize(+0x2908) [0xaaaadff02908]
Bug: Assertion (disk != NULL) at ../../libparted/disk.c:1620 in function ped_disk_get_partition_by_sector() failed.
Aborted

It is able to read the file:

parallels@debian-gnu-linux-11:~/fatresize$ /usr/sbin/fatresize -i  boot_partition.fs
fatresize 1.1.0 (20201114)
part(start=0, end=524287, length=524288)
FAT: fat32
Cur size: 268435456
Min size: 44493312
Max size: 268435456

Fatresize does nothing except print version

When trying to shrink a fat16 partition, fatresize just spits out its version and does nothing.

[james@voidD620 ~]$ lsblk
NAME                MAJ:MIN RM  SIZE RO TYPE  MOUNTPOINTS
sda                   8:0    0  1.8T  0 disk  
└─sda1                8:1    0  1.8T  0 part  
  └─voidD620        254:0    0  1.8T  0 crypt 
    └─voidD620-root 254:1    0  1.8T  0 lvm   /
sdb                   8:16   1 59.1G  0 disk  
├─sdb1                8:17   1   59G  0 part  
└─sdb2                8:18   1   32M  0 part  
sr0                  11:0    1 1024M  0 rom   
[james@voidD620 ~]$ sudo fatresize -s 28M /dev/sdb2
Password: 
fatresize 1.0.3 (20200810)
[james@voidD620 ~]$ lsblk
NAME                MAJ:MIN RM  SIZE RO TYPE  MOUNTPOINTS
sda                   8:0    0  1.8T  0 disk  
└─sda1                8:1    0  1.8T  0 part  
  └─voidD620        254:0    0  1.8T  0 crypt 
    └─voidD620-root 254:1    0  1.8T  0 lvm   /
sdb                   8:16   1 59.1G  0 disk  
├─sdb1                8:17   1   59G  0 part  
└─sdb2                8:18   1   32M  0 part  
sr0                  11:0    1 1024M  0 rom   
[james@voidD620 ~]$

Disk /dev/sdb: 59.06 GiB, 63417876480 bytes, 123863040 sectors
Disk model: USB 3.0 FD      
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: 0x1c9d51dc

Device     Boot     Start       End   Sectors Size Id Type
/dev/sdb1  *         2048 123797503 123795456  59G  7 HPFS/NTFS/exFAT
/dev/sdb2       123797504 123863039     65536  32M ef EFI (FAT-12/16/32)

ESP, a FAT32 FS, rejected

My EFI System Partition is too small (100m; Microsoft's fault). So I'm trying to grow it, with mixed results.

gparted will grow the partition but not the filesystem.
fatresize doesn't recognize the FS.

$ sudo fatresize -i /dev/sda2
fails with the message: Error: /dev/sda2 is ntfs, not a valid FAT16/FAT32 partition.

(The wording has changed due to a commit (I just issued a pull request for it.))

So: fatresize thinks that the FS (or, perhaps, the partition type) is ntfs. Why?

parted does not think that this is an NTFS. It says:

Number Start End Size File system Name Flags
2 524MB 733MB 209MB fat32 EFI System Partition boot, esp

Why does fatresize thinks that this partition is ntfs?
Any recommendations on how to proceed?

(I don't know if there are fancy things about an ESP that I should beware of.)

Strange error for loop devices like /dev/loop0: "You must specify exactly one existing device."

As I don't have a physical device for my floppy image (see #14), I've created a /dev/loop0 device with a command

sudo losetup /dev/loop1 ./floppy.img

However, when I try to use a command like

sudo fatresize -s 2880k /dev/loop1

I get this output:

fatresize 1.0.3 (03/05/20)
You must specify exactly one existing device.

Perhaps this has to do with /dev/loop1 device having one more letter than i.e. /dev/sdd1

./configure gives a "bad version string" for libparted 3.3 because it's not a "1.6.0" format

Temporary solution is to insert a return 0; to line 4795 to skip this check.

Error to build 1.0.4

I'm an ArchLinux user, when I try to compile version 1.0.4 the compilation process is interrupted with the following error:

make[1]: Entering directory '/tmp/makepkg/fatresize/src/fatresize-1.0.4'
gcc  -march=x86-64 -mtune=generic -O2 -pipe -fstack-protector-strong -fno-plt -D_FILE_OFFSET_BITS=no  -Wl,-O1,--sort-common,--as-needed,-z,relro,-z,now -o fatresize fatresize.o -lparted-fs-resize -lparted -ldl
fatresize.o: In function `try_snap':
fatresize.c:(.text.unlikely+0xad): undefined reference to `ped_assert'
fatresize.o: In function `main':
fatresize.c:(.text.startup+0xc83): undefined reference to `ped_assert'
fatresize.c:(.text.startup+0xd42): undefined reference to `ped_assert'
collect2: error: ld returned 1 exit status
make[1]: *** [Makefile:416: fatresize] Error 1
make[1]: Leaving directory '/tmp/makepkg/fatresize/src/fatresize-1.0.4'
make: *** [Makefile:319: all] Error 2
==> ERROR: There was an error to build ().
     Canceling ...

Any idea or suggestion?
Thanks

fatresize should force user confirmation before performing irreversible and unwanted changes

In the following case I wished to resize a 512MiB FAT16 filesystem to a 4GiB one but forgot about the 2GiB limit. This is a log:

# fatresize -s 3700Mi -i /dev/sda1
fatresize 1.0.2 (06/16/18)
FAT: fat16
Size: 512451072
Min size: 489846784
Max size: 4000317440

# fatresize -s 3700Mi /dev/sda1
fatresize 1.0.2 (06/16/18)
Warning: The file system can only be resized to this size by converting to FAT32. 
If you convert to FAT32, and MS Windows is installed on this partition, then you
must re-install the MS Windows boot loader.  If you want to do this, you should
consult the Parted manual (or your distribution's manual).  Also, converting to
FAT32 will make the file system unreadable by MS DOS, MS Windows 95a, and
MS Windows NT.

# fatresize -i /dev/sda1
fatresize 1.0.2 (06/16/18)
FAT: fat32
Size: 3879977472
Min size: 536870912
Max size: 4000317440

fatresize provided a warning but did not ask me to confirm that I wanted to proceed before converting the filesystem to FAT32, which (1) is surprising behavior, (2) is not something I asked the tool to do, (3) left the filesystem in a useless state since this was being used for a DOS machine.

My suggestion is to implement one of the following:

Prompt the user to type the word "yes" to confirm that filesystem conversion is wanted.
Change the warning to an error and force the user to supply a command line argument like "--allow-filesystem-conversion".

manpage broken on arch linux

Probably something to do with docbook-to-man

FAT16 resize error

I have cloned the 1.1.0 but the binary shows 1.0.3 with the date of compilation

roberto@vm-ubuntu18:~/tinycore-editor$ sudo fatresize -i /dev/loop0p1
fatresize 1.0.3 (20210815)
part(start=2048, end=266239, length=264192)
FAT: fat16
Cur size: 67108864
Min size: 33728000
Max size: 136314880

roberto@vm-ubuntu18:~/tinycore-editor$ sudo fatresize -s 136314880 /dev/loop0p1
fatresize 1.0.3 (20210815)
part(start=2048, end=266239, length=264192)
Error: The location 137MB is outside of the device /dev/loop0.

roberto@vm-ubuntu18:~/tinycore-editor$ sudo fatresize -s 129M /dev/loop0p1
fatresize 1.0.3 (20210815)
part(start=2048, end=266239, length=264192)
No Implementation: GNU Parted cannot resize this partition to this size. We're working on it!

"Bad macro" build errors

Probably happen because of incompatibility with a modern libparted.

fatresize.c: In function ‘snap’:
fatresize.c:268:74: error: macro "PED_ASSERT" passed 2 arguments, but takes just 1
     FAT_ASSERT(ped_geometry_test_sector_inside (range, *sector), return 0);
                                                                          ^
fatresize.c:51:34: error: ‘PED_ASSERT’ undeclared (first use in this function)
 #define FAT_ASSERT(cond, action) PED_ASSERT(cond, action)
                                  ^
fatresize.c:268:5: note: in expansion of macro ‘FAT_ASSERT’
     FAT_ASSERT(ped_geometry_test_sector_inside (range, *sector), return 0);
     ^~~~~~~~~~
fatresize.c:51:34: note: each undeclared identifier is reported only once for each function it appears in
 #define FAT_ASSERT(cond, action) PED_ASSERT(cond, action)
                                  ^
fatresize.c:268:5: note: in expansion of macro ‘FAT_ASSERT’
     FAT_ASSERT(ped_geometry_test_sector_inside (range, *sector), return 0);
     ^~~~~~~~~~
fatresize.c: In function ‘snap_to_boundaries’:
fatresize.c:317:32: error: macro "PED_ASSERT" passed 2 arguments, but takes just 1
  FAT_ASSERT (start_part, return);
                                ^
fatresize.c:51:34: error: ‘PED_ASSERT’ undeclared (first use in this function)
 #define FAT_ASSERT(cond, action) PED_ASSERT(cond, action)
                                  ^
fatresize.c:317:2: note: in expansion of macro ‘FAT_ASSERT’
  FAT_ASSERT (start_part, return);
  ^~~~~~~~~~
fatresize.c:337:34: error: macro "PED_ASSERT" passed 2 arguments, but takes just 1
  FAT_ASSERT (start <= end, return);

"Killed" with no verbosity

Attempted on an external WD:

~ $ sudo fatresize -p -v -s 1800G /dev/sdb2
fatresize 1.0.2 (06/16/18)
.Resizing file system.
Killed

Option for working with fat32 partitions less than 512 MiB

It would be great, if fatresize had a special option for working with fat32 partitions less than 512 MiB.

Mostly many EFI partitions has size less than 512 MiB and fatresize can't work properly with them.

The most incomprehensible thing is why "fatresize -i" show "Min size" also limited to 512 MiB.