std: micro-optimize thread local accesses

[joboet]

Alternative to #143511.

I had some further ideas on how to optimise TLS accesses...

[joboet]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 45cb56f (45cb56f797a0bb4c02cd34f99c3fe416b3065c38, parent: 8954863c81df429ebf96ea38a16c76f209995833)

[rust-timer]

Finished benchmarking commit (45cb56f): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

@rustbot label: -S-waiting-on-perf -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.2%	[-0.2%, -0.2%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 2.3%, secondary 3.6%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	6.6%	[6.1%, 7.6%]	3
Regressions ❌ (secondary)	3.6%	[2.0%, 6.5%]	3
Improvements ✅ (primary)	-4.2%	[-5.4%, -3.0%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.3%	[-5.4%, 7.6%]	5

Cycles

Results (primary 3.4%, secondary 3.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	3.4%	[1.9%, 6.7%]	101
Regressions ❌ (secondary)	3.4%	[1.8%, 7.7%]	95
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.1%	[-2.1%, -2.1%]	1
All ❌✅ (primary)	3.4%	[1.9%, 6.7%]	101

Binary size

Results (primary 0.0%, secondary 0.0%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	0.1%	[0.0%, 0.4%]	44
Regressions ❌ (secondary)	0.0%	[0.0%, 0.1%]	38
Improvements ✅ (primary)	-0.4%	[-1.0%, -0.1%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.0%	[-1.0%, 0.4%]	47

Bootstrap: 514.086s -> 515.048s (0.19%)
Artifact size: 400.74 MiB -> 400.69 MiB (-0.01%)

[joboet]

Cycle times appear to be way worse, so this doesn't work out apparently...