fix(inode): use atomic.Int64 for prevDirListingTimeStamp to fix lock-free data race

[kislaykishore]

Description

This PR refactors dirInode.prevDirListingTimeStamp from a 24-byte non-atomic time.Time struct to an 8-byte atomic.Int64 (storing unix nanoseconds).

Motivation & Problem Statement

ShouldInvalidateKernelListCache(), InvalidateKernelListCache(), and ReadEntryCores() accessed prevDirListingTimeStamp concurrently without synchronization locks. Because time.Time is a 24-byte 3-word composite struct (wall, ext, loc), concurrent non-atomic reads and writes constituted a multi-word data race on 64-bit architectures.
Replacing time.Time with atomic.Int64 provides lock-free hardware atomics (Load() / Store()), completely eliminating the data race while retaining lock-free cache invalidation performance.

Struct Sizing & Performance Impact

• Shrinks dirInode by 16 bytes (240 B -> 224 B).
• CPU Throughput: Evaluates in ~0.75 ns/op via lock-free atomic load.
• Race Verification: Verified with 0 data races under go test -race across 128 parallel goroutines.

Benchmark Target	Master Baseline	Optimized Branch	CPU Latency Delta	Concurrency & Race Status (-race)
ShouldInvalidateKernelListCache_Parallel	0.803 ns/op	0.756 ns/op	-5.85% ($p=0.029$)	0 races (Lock-free atomic load)
ShouldInvalidateKernelListCache_Expired	42.79 ns/op	42.12 ns/op	-1.55% ($p=0.000$)	0 races
ShouldInvalidateKernelListCache_NotListed	2.043 ns/op	1.918 ns/op	-6.14% ($p=0.000$)	0 races (Fast-path zero check)

Link to the issue in case of a bug fix.

N/A

Testing details

1. Manual - Executed make build and verified formatting/linter with 0 issues.
2. Unit tests - Executed unit tests and concurrency race tests: go test -v -race ./internal/fs/inode/....
3. Integration tests - N/A

Any backward incompatible change? If so, please explain.

N/A

[gemini-code-assist]

Code Review

This pull request refactors prevDirListingTimeStamp in dirInode from time.Time to atomic.Int64 to resolve concurrency data races during lock-free reads and concurrent writes. Corresponding updates are made to ReadEntryCores, ShouldInvalidateKernelListCache, and InvalidateKernelListCache to use atomic operations, and a new concurrency test is added to verify the fix. There are no review comments, so I have no feedback to provide.

[codecov]

Codecov Report

❌ Patch coverage is 85.71429% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 83.82%. Comparing base (0673d19) to head (63897ab).

Files with missing lines	Patch %	Lines
internal/fs/inode/dir.go	85.71%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #4999   +/-   ##
=======================================
  Coverage   83.81%   83.82%           
=======================================
  Files         174      174           
  Lines       21376    21385    +9     
=======================================
+ Hits        17916    17925    +9     
  Misses       2779     2779           
  Partials      681      681           

Flag	Coverage Δ
unittests	83.82% <85.71%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.