Add AggressiveInlining to hex conversion methods

[xtqqczze]

Context: #124702 (comment)

cc: @EgorBo

[xtqqczze]

@MihuBot

[dotnet-policy-service]

Tagging subscribers to this area: @karelz, @dotnet/ncl
See info in area-owners.md if you want to be subscribed.

[MihaZupan]

@MihuBot benchmark Perf_Uri -medium

[MihuBot]

System.Tests.Perf_Uri

BenchmarkDotNet v0.16.0-nightly.20260320.467, Linux Ubuntu 24.04.4 LTS (Noble Numbat)
AMD EPYC 9V74 2.60GHz, 1 CPU, 8 logical and 4 physical cores
Memory: 31.34 GB Total, 5.38 GB Available
MediumRun : .NET 11.0.0 (11.0.0-dev, 42.42.42.42424), X64 RyuJIT x86-64-v4
Job=MediumRun  IterationCount=15  LaunchCount=2
WarmupCount=10

Method	Toolchain	input	Mean	Error	Ratio	Allocated	Alloc Ratio
ParseAbsoluteUri	Main	?	166.112 ns	0.7540 ns	1.00	304 B	1.00
ParseAbsoluteUri	PR	?	168.810 ns	2.2408 ns	1.02	304 B	1.00
DnsSafeHost	Main	?	154.589 ns	1.2504 ns	1.00	328 B	1.00
DnsSafeHost	PR	?	149.468 ns	0.4216 ns	0.97	328 B	1.00
BuilderToString	Main	?	72.996 ns	0.1749 ns	1.00	216 B	1.00
BuilderToString	PR	?	73.671 ns	0.2681 ns	1.01	216 B	1.00
UriBuilderReplacePort	Main	?	69.379 ns	0.3782 ns	1.00	216 B	1.00
UriBuilderReplacePort	PR	?	73.594 ns	2.7388 ns	1.06	216 B	1.00
GetComponents	Main	?	12.314 ns	0.2345 ns	1.00	80 B	1.00
GetComponents	PR	?	12.000 ns	0.1246 ns	0.98	80 B	1.00
PathAndQuery	Main	?	1.950 ns	0.0093 ns	1.00	-	NA
PathAndQuery	PR	?	1.938 ns	0.0010 ns	0.99	-	NA
EscapeDataString	Main	{{{{{{{{{{{{(...){{{{{{{{{{{{ [1000]	4,749.169 ns	17.0779 ns	1.00	6024 B	1.00
EscapeDataString	PR	{{{{{{{{{{{{(...){{{{{{{{{{{{ [1000]	4,776.033 ns	25.3911 ns	1.01	6024 B	1.00
CombineAbsoluteRelative	Main	/new/path	91.880 ns	0.1661 ns	1.00	200 B	1.00
CombineAbsoluteRelative	PR	/new/path	90.979 ns	0.2487 ns	0.99	200 B	1.00
UnescapeDataString	Main	%E4%BD%A0%E5%A5%BD	40.370 ns	0.0832 ns	1.00	32 B	1.00
UnescapeDataString	PR	%E4%BD%A0%E5%A5%BD	40.498 ns	0.0567 ns	1.00	32 B	1.00
EscapeDataString	Main	a{üa{üa{üa{ü(...)a{üa{üa{üa{ü [999]	7,722.241 ns	68.2582 ns	1.00	6688 B	1.00
EscapeDataString	PR	a{üa{üa{üa{ü(...)a{üa{üa{üa{ü [999]	7,690.173 ns	26.1426 ns	1.00	6688 B	1.00
EscapeDataString	Main	aaaaaaaaaaaa(...)aaaaaaaaaaaa [1000]	27.709 ns	0.0162 ns	1.00	-	NA
EscapeDataString	PR	aaaaaaaaaaaa(...)aaaaaaaaaaaa [1000]	27.967 ns	0.0153 ns	1.01	-	NA
UnescapeDataString	Main	abc%20def%20ghi%20	31.766 ns	0.0838 ns	1.00	48 B	1.00
UnescapeDataString	PR	abc%20def%20ghi%20	31.500 ns	0.0579 ns	0.99	48 B	1.00
Ctor	Main	http://dot.net	53.408 ns	0.0801 ns	1.00	56 B	1.00
Ctor	PR	http://dot.net	56.051 ns	1.6965 ns	1.05	56 B	1.00
CtorIdnHostPathAndQuery	Main	http://dot.ne(...)alue#fragment [43]	159.449 ns	0.5115 ns	1.00	248 B	1.00
CtorIdnHostPathAndQuery	PR	http://dot.ne(...)alue#fragment [43]	158.283 ns	0.2227 ns	0.99	248 B	1.00
Ctor	Main	http://höst.with.ünicode	217.115 ns	2.2451 ns	1.00	328 B	1.00
Ctor	PR	http://höst.with.ünicode	226.128 ns	1.9344 ns	1.04	328 B	1.00
CtorIdnHostPathAndQuery	Main	http://höst.w(...)alue#fragment [53]	930.307 ns	2.9512 ns	1.00	592 B	1.00
CtorIdnHostPathAndQuery	PR	http://höst.w(...)alue#fragment [53]	965.755 ns	13.8810 ns	1.04	592 B	1.00
CtorIdnHostPathAndQuery	Main	http://host/	113.443 ns	0.4717 ns	1.00	200 B	1.00
CtorIdnHostPathAndQuery	PR	http://host/	112.996 ns	0.3394 ns	1.00	200 B	1.00
CtorIdnHostPathAndQuery	Main	http://host/p(...)s?key=ünicode [50]	383.232 ns	1.5703 ns	1.00	520 B	1.00
CtorIdnHostPathAndQuery	PR	http://host/p(...)s?key=ünicode [50]	383.020 ns	2.0580 ns	1.00	520 B	1.00
CtorIdnHostPathAndQuery	Main	http://host/p(...)es?key=va lue [49]	247.907 ns	1.0858 ns	1.00	296 B	1.00
CtorIdnHostPathAndQuery	PR	http://host/p(...)es?key=va lue [49]	253.543 ns	1.1261 ns	1.02	296 B	1.00
CtorIdnHostPathAndQuery	Main	http://host/p(...)3&key4=value4 [64]	179.891 ns	0.4750 ns	1.00	304 B	1.00
CtorIdnHostPathAndQuery	PR	http://host/p(...)3&key4=value4 [64]	177.851 ns	0.8947 ns	0.99	304 B	1.00
CtorIdnHostPathAndQuery	Main	http://host/p(...)=%C3%BCnicode [61]	413.036 ns	3.0807 ns	1.00	520 B	1.00
CtorIdnHostPathAndQuery	PR	http://host/p(...)=%C3%BCnicode [61]	435.076 ns	1.4686 ns	1.05	520 B	1.00
CtorIdnHostPathAndQuery	Main	http://host/p(...)?key=va%20lue [57]	196.776 ns	1.3725 ns	1.00	296 B	1.00
CtorIdnHostPathAndQuery	PR	http://host/p(...)?key=va%20lue [57]	196.296 ns	1.3107 ns	1.00	296 B	1.00
Ctor	Main	http://xn--hs(...)n--nicode-2ya [38]	52.883 ns	0.0579 ns	1.00	56 B	1.00
Ctor	PR	http://xn--hs(...)n--nicode-2ya [38]	52.919 ns	0.0908 ns	1.00	56 B	1.00
CtorIdnHostPathAndQuery	Main	http://xn--hs(...)alue#fragment [67]	187.568 ns	0.9054 ns	1.00	296 B	1.00
CtorIdnHostPathAndQuery	PR	http://xn--hs(...)alue#fragment [67]	187.150 ns	2.0227 ns	1.00	296 B	1.00
Ctor	Main	https://a.much.longer.domain.name	67.964 ns	0.0924 ns	1.00	56 B	1.00
Ctor	PR	https://a.much.longer.domain.name	67.864 ns	0.0536 ns	1.00	56 B	1.00
CtorIdnHostPathAndQuery	Main	https://a.muc(...)alue#fragment [62]	193.176 ns	0.2980 ns	1.00	280 B	1.00
CtorIdnHostPathAndQuery	PR	https://a.muc(...)alue#fragment [62]	206.270 ns	4.8661 ns	1.07	280 B	1.00
Ctor	Main	https://contoso.com	45.542 ns	0.0494 ns	1.00	56 B	1.00
Ctor	PR	https://contoso.com	45.530 ns	0.0611 ns	1.00	56 B	1.00
Ctor	Main	https://CONTOSO.com	45.593 ns	0.0527 ns	1.00	56 B	1.00
Ctor	PR	https://CONTOSO.com	45.728 ns	0.0707 ns	1.00	56 B	1.00
CtorIdnHostPathAndQuery	Main	https://conto(...)alue#fragment [48]	158.171 ns	0.2208 ns	1.00	256 B	1.00
CtorIdnHostPathAndQuery	PR	https://conto(...)alue#fragment [48]	159.140 ns	0.4518 ns	1.01	256 B	1.00
CtorIdnHostPathAndQuery	Main	https://CONTO(...)alue#fragment [48]	161.100 ns	0.9257 ns	1.00	256 B	1.00
CtorIdnHostPathAndQuery	PR	https://CONTO(...)alue#fragment [48]	166.534 ns	3.0766 ns	1.03	256 B	1.00
EscapeDataString	Main	üüüüüüüüüüüü(...)üüüüüüüüüüüü [1000]	8,922.806 ns	64.6769 ns	1.00	12024 B	1.00
EscapeDataString	PR	üüüüüüüüüüüü(...)üüüüüüüüüüüü [1000]	8,884.934 ns	32.0379 ns	1.00	12024 B	1.00

[EgorBo]

Is there a scenario it improves? It doesn't seem to improve these benchmarks.
AggressiveInlining should be backed by real benchmarks. E.g. in "reduce unsafe" work we sometimes use it to patch potential regressions after moving to safe code (and it's driven by benchmarks).