fix: race condition for device error propagation

[madjesc]

fix: file backend NO_SENSE errortype when it should be backwards
fix: unified simple problems

Changelog

Fixed

Race Condition in Device Error Propagation

• Resolved race condition in error propagation mechanism for device operations
• Centralized write error handling in iosched.c to prevent inconsistent error states across multiple code paths

Error Type Correction in File Debug Driver

• Fixed incorrect error type mapping in filedebug_tc.c where NO_SENSE and WRITE_PERM error types were reversed
• Now correctly returns EDEV_WRITE_PERM when force_errortype is set, and EDEV_NO_SENSE otherwise

Write Permission Error Handling

• Improved handling of write permission errors by consolidating logic in I/O scheduler layer
• Removed redundant MAM (Medium Auxiliary Memory) volume lock updates from multiple locations
• Centralized MAM lock status updates to prevent race conditions during error scenarios

Changed

Code Refactoring

• Moved _unified_write_index_after_perm function from unified.c to iosched.c as _iosched_write_index_after_perm
• Removed duplicate error handling code from unified scheduler
• Simplified error return logic in ltfs_fsops_write function

Error Handling Architecture

• Consolidated write error handling at the I/O scheduler level
• Removed scattered MAM volume lock updates from ltfs_write_index function
• Improved error propagation flow from low-level write operations to higher layers

Added

New Error Checking Macro

• Added IS_RW_PERM macro in ltfs_error.h for checking read/write permission errors

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have confirmed my fix is effective or that my feature works

[amissael95]

Thanks for you efforts and working in this change. Leaving a couple of comments. Please ensure the fix is tested and documented in this PR properly.

[amissael95]

Why are you removing these lines? Looks like you do not longer want to set the volstat value, do you plan to set this values from somewhere else?

[Piloalucard]

Is this header necessary? I see deletions only in this file

[Piloalucard]

Is this the best way to obtain the Partition to lock?
I also see we do not mantain the PWE_MAM_BOTH logic in case a PWE is already present in a partition, this could not lead to problems on analysis?