shell: modules: devmem: streamline the codebase #98437
Conversation
To improve code clarity, use `shell_print` in place of `shell_fprintf` with `SHELL_NORMAL` where appropriate. Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
| hex_data[data_offset] = (uint8_t)value; | ||
| value >>= 8; | ||
| hex_data[data_offset + 1] = (uint8_t)value; | ||
| sys_put_be16(value, &hex_data[data_offset]); |
There was a problem hiding this comment.
static inline void sys_put_be16(uint16_t val, uint8_t dst[2])
{
dst[0] = val >> 8;
dst[1] = val;
}You probably should use sys_put_le16() to maintain the same endianness?
static inline void sys_put_le16(uint16_t val, uint8_t dst[2])
{
dst[0] = val;
dst[1] = val >> 8;
}There was a problem hiding this comment.
Sorry for my misinterpretation. It seems that devmem dump will always output memory in Little-Endian byte order, regardless of the underlying architecture. The correct one is as follows:
case 16:
value = sys_read16(addr + data_offset);
sys_put_le16(value, &hex_data[data_offset]);
break;
case 32:
value = sys_read32(addr + data_offset);
sys_put_le32(value, &hex_data[data_offset]);
break;
#ifdef CONFIG_64BIT
case 64:
value = sys_read64(addr + data_offset);
sys_put_le64(value, &hex_data[data_offset]);
break;
#endif /* CONFIG_64BIT */
| break; | ||
| case 16: | ||
| value = sys_le16_to_cpu(sys_read16(addr + data_offset)); | ||
| value = sys_read16(addr + data_offset); |
There was a problem hiding this comment.
Maybe you should add big endian support for this shell command instead of converting it to big endian?
There was a problem hiding this comment.
Maybe you should add big endian support for this shell command instead of converting it to big endian?
Thanks for the suggestion! For now, I'm focusing on streamlining the existing codebase, so I'll hold off on adding big-endian support in this PR.
Replaces direct byte shifting and assignment with `sys_put_le16`, `sys_put_le32`, and `sys_put_le64` to simplify the `memory_dump` function. Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
ndrs-pst
force-pushed
the
pr_shell_devmem_service_2025-10-28
branch
from
bf9176d to
44a2a2b
Compare
|
|
Ping @ycsin. |
5 participants
This PR is dedicated to improving
devmem_service.cthrough the following commits:shell_printwhere applicablesys_put_le(16|32|64)inmemory_dump