libs/libc: Fix divide-by-zero in stat() with large filesystem block sizes

[ML-dev-crypto]

Summary

This change fixes a divide-by-zero in stat() caused by overflow of st_blksize when a filesystem reports a block size larger than can be represented by blksize_t. As reported in #19432, this can occur with LittleFS configurations using large block sizes, where blksize_t is currently defined as int16_t. When the filesystem block size exceeds the range of int16_t, st_blksize is truncated to zero, causing an integer divide-by-zero when st_blocks is calculated.

This patch widens blksize_t from int16_t to int32_t in include/sys/types.h and updates the corresponding nuttx_blksize_t definition in include/nuttx/fs/hostfs.h so the two definitions remain consistent.

Fixes #19432

---

Impact

This change allows st_blksize to correctly represent larger filesystem block sizes and prevents the divide-by-zero reported in #19432. It does not introduce any new configuration options or dependencies, and only updates the blksize_t typedef together with the matching nuttx_blksize_t definition used by hostfs.

---

Testing

Host: Ubuntu 24.04 (Docker), x86_64

The simulator was configured with CONFIG_RAMMTD_ERASESIZE=4096 and CONFIG_FS_LITTLEFS_BLOCK_SIZE_FACTOR=16, resulting in a LittleFS block size of 65536 bytes. The project built successfully using:

make -j$(nproc)

which produced:

LD:  nuttx
Pac SIM with dynamic libs..
'/lib/x86_64-linux-gnu/libz.so.1' -> 'sim-pac/libs/libz.so.1'
'/lib/x86_64-linux-gnu/libc.so.6' -> 'sim-pac/libs/libc.so.6'
'/lib64/ld-linux-x86-64.so.2' -> 'sim-pac/ld-linux-x86-64.so.2'
SIM elf with dynamic libs archive in nuttx.tgz

After booting the simulator, LittleFS was mounted successfully. A file was created and listed successfully using ls -l, which exercises the stat() path involved in the reported issue.

nsh> mount
  /mnt/lfs type littlefs

nsh> echo "test" > /mnt/lfs/testfile

nsh> ls -l /mnt/lfs
/mnt/lfs:
 drwxrwxrwx           0 .
 drwxrwxrwx           0 ..
 -rw-rw-rw-           5 2026-07-15 17:24 testfile

nsh> ls -l /mnt/lfs/testfile
-rw-rw-rw-           5 2026-07-15 17:24 /mnt/lfs/testfile

With this configuration, no divide-by-zero or simulator crash was observed while listing files on the LittleFS mount.

[linguini1]

Please follow the contribution guide and PR template :)
Also, please include some logs from your tests of this change.

[ML-dev-crypto]

Thanks! I'll update the PR to follow the template and rerun the tests locally so I can include the requested build and test logs.

[github-actions]

MemBrowse Memory Report

esp32-devkitc

• ROM: .flash.text -8 B (-0.0%, 122,708 B / 4,194,272 B, total: 3% used)
• irom0_0_seg: .flash.text -8 B (-0.0%, 87,076 B / 3,342,304 B, total: 3% used)

hifive1-revb

• flash: .text -12 B (-0.0%, 83,240 B / 4,194,304 B, total: 2% used)

qemu-armv8a

• Code: .text.arm64_fpu_procfs_stat -4 B, .text.file_vioctl -4 B, .text.inode_stat -4 B (-0.0%, 315,362 B)

qemu-intel64

• Code: .text -11 B (-0.0%, 8,655,434 B)

stm32-nucleo-f103rb

• flash: .text -4 B (-0.0%, 33,788 B / 131,072 B, total: 26% used)
  No memory changes detected for:
• arduino-mega2560
• mirtoo
• rx65n-rsk2mb
• s698pm-dkit

[linguini1]

It looks like you will need to fix the log statements that expect the block size to be 16-bit