') + ')', 'gi'); if (regex.test(text)) { found = true; var frag = document.createDocumentFragment(); var parts = text.split(regex); parts.forEach(function(part, i) { if (i % 2 === 0) { frag.appendChild(document.createTextNode(part)); } else { var span = document.createElement('span'); span.className = 'userscript-highlight'; span.textContent = part; frag.appendChild(span); } }); node.parentNode.replaceChild(frag, node); } }); } else if (node.nodeType === 1 && node.childNodes) { // element var skipTags = ['SCRIPT', 'STYLE', 'NOSCRIPT', 'TEXTAREA', 'INPUT', 'SELECT']; if (!skipTags.includes(node.tagName)) { Array.from(node.childNodes).forEach(highlight); } } } highlight(document.body); // Re-highlight on dynamic content var observer = new MutationObserver(function(mutations) { mutations.forEach(function(m) { m.addedNodes.forEach(function(node) { if (node.nodeType === 1 || node.nodeType === 3) highlight(node); }); }); }); observer.observe(document.body, { childList: true, subtree: true }); })(); } } catch(__e) { console.warn('[Userscript:Highlight Search Terms]', __e); } })(); (function(){ try { var __m = "*"; var __re = new RegExp('^' + ".*" + ', 'i'); if (__m === '*' || __re.test(location.href)) { // Strip utm_, fbclid, gclid, etc. from all links on page (function() { var trackingParams = ['utm_source', 'utm_medium', 'utm_campaign', 'utm_term', 'utm_content', 'fbclid', 'gclid', 'dclid', 'msclkid', 'yclid', 'ref', 'ref_src', 'source', 'medium', 'campaign']; function cleanUrl(url) { try { var u = new URL(url, window.location.origin); var changed = false; trackingParams.forEach(function(p) { if (u.searchParams.has(p)) { u.searchParams.delete(p); changed = true; } }); return changed ? u.toString() : url; } catch (e) { return url; } } function cleanLinks() { document.querySelectorAll('a[href]').forEach(function(a) { var clean = cleanUrl(a.href); if (clean !== a.href) a.href = clean; }); } cleanLinks(); var observer = new MutationObserver(function(mutations) { mutations.forEach(function(m) { m.addedNodes.forEach(function(node) { if (node.nodeType === 1) { if (node.tagName === 'A') cleanLinks(); node.querySelectorAll('a[href]').forEach(function(a) { var clean = cleanUrl(a.href); if (clean !== a.href) a.href = clean; }); } }); }); }); observer.observe(document.body, { childList: true, subtree: true }); })(); } } catch(__e) { console.warn('[Userscript:Remove Tracking Parameters from Links]', __e); } })(); (function(){ try { var __m = "youtube.com"; var __re = new RegExp('^' + "youtube\\.com" + ', 'i'); if (__m === '*' || __re.test(location.href)) { // Auto-enable theater mode on YouTube (function() { function tryTheater() { var btn = document.querySelector('button[aria-label="Theater mode"], ytd-player #player button[title="Theater mode"]'); if (btn && !btn.classList.contains('activated')) { btn.click(); } } // Try immediately tryTheater(); // Try after navigation (SPA) var lastUrl = location.href; setInterval(function() { if (location.href !== lastUrl) { lastUrl = location.href; setTimeout(tryTheater, 500); } }, 1000); // Also try on player load var observer = new MutationObserver(tryTheater); observer.observe(document.body, { childList: true, subtree: true }); })(); } } catch(__e) { console.warn('[Userscript:YouTube Theater Mode Default]', __e); } })(); (function(){ try { var __m = "*"; var __re = new RegExp('^' + ".*" + ', 'i'); if (__m === '*' || __re.test(location.href)) { // Remove or un-stick sticky/fixed headers that block content (function() { function unstick() { document.querySelectorAll('header, nav, [role="banner"], .header, .navbar, .sticky, .fixed-top, [style*="position: fixed"], [style*="position:sticky"]').forEach(function(el) { if (el.style.position === 'fixed' || el.style.position === 'sticky' || getComputedStyle(el).position === 'fixed' || getComputedStyle(el).position === 'sticky') { el.style.position = 'static'; el.style.top = 'auto'; el.style.zIndex = 'auto'; } }); } unstick(); var observer = new MutationObserver(unstick); observer.observe(document.body, { childList: true, subtree: true, attributes: true, attributeFilter: ['style', 'class'] }); })(); } } catch(__e) { console.warn('[Userscript:Kill Sticky Headers]', __e); } })(); })(); Inline TLS field access for linux/osx x64/arm64 by kunalspathak · Pull Request #87082 · dotnet/runtime · GitHub
Skip to content

Inline TLS field access for linux/osx x64/arm64 - #87082

Merged
kunalspathak merged 88 commits into
dotnet:mainfrom
kunalspathak:tls_linux
Jul 6, 2023
Merged

Inline TLS field access for linux/osx x64/arm64#87082
kunalspathak merged 88 commits into
dotnet:mainfrom
kunalspathak:tls_linux

Conversation

@kunalspathak

@kunalspathakkunalspathak commented Jun 2, 2023•
edited
Loading

Copy link
Copy Markdown
Contributor

Extending #85619 and #82973 for linux.

movrdi,0x7FA9DB182F28 ; GOT descriptor address of `maxThreadStaticBlocks`movrax,0x7FA9DB7A5DB0 ; function address of __tls_get_addrcallrax ; gets the address of `maxThreadStaticBlocks`cmp dword ptr [rax],2jl SHORT G_M8151_IG06movrdi, qword ptr [rax+08H] ; access the `threadStaticBlocks` variablemovrdi, qword ptr [rdi+10H]testrdi,rdije SHORT G_M8151_IG06movrax, bword ptr [rdi]addrax,16

Base:

| Method | Mean | Error | StdDev | Median | Min | Max | Allocated |
|---------------- |---------:|----------:|----------:|---------:|---------:|---------:|----------:|
| GetThreadStatic | 3.383 ns | 0.0378 ns | 0.0316 ns | 3.374 ns | 3.346 ns | 3.459 ns | - |
| SetThreadStatic | 5.334 ns | 0.0733 ns | 0.0650 ns | 5.338 ns | 5.261 ns | 5.451 ns | - |

Diff:

| Method | Mean | Error | StdDev | Median | Min | Max | Allocated |
|---------------- |---------:|----------:|----------:|---------:|---------:|---------:|----------:|
| GetThreadStatic | 1.048 ns | 0.0200 ns | 0.0177 ns | 1.041 ns | 1.026 ns | 1.087 ns | - |
| SetThreadStatic | 2.844 ns | 0.0379 ns | 0.0354 ns | 2.837 ns | 2.807 ns | 2.915 ns | - |

This also adds mrs instruction in arm64 to access TLS on linux. The codegen is:

G_M8151_IG02:  mrs x0, tpidr_el0 ldr w1,[x0, #0xE8] ; `maxThreadStaticBlocks` valuecmp w1, #2 blt G_M8151_IG05 ldr x0,[x0, #0xF0] ; `threadStaticBlocks` array ldr x1,[x0, #0x10] cbz x1, G_M8151_IG05G_M8151_IG03: mov w0, #5str w0,[x1, #0x24]

osx/arm64:

Because of register allocator doesn't reuse the x0 properly, we end up creating the same constant again in x1, but this is how osx/arm64 looks like:

 movz x0, #0x8C00 movk x0, #817LSL #16 movk x0, #1LSL #32 movz x1, #0x8C00 movk x1, #817LSL #16 movk x1, #1LSL #32 ldr x1,[x1] blr x1

osx/x64:

image

@ghostghost added the area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI label Jun 2, 2023
@ghost

ghost commented Jun 2, 2023

Copy link
Copy Markdown

Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch
See info in area-owners.md if you want to be subscribed.

Issue Details

Extending #85619 and #82973 for linux.

Author:kunalspathak
Assignees:-
Labels:

area-CodeGen-coreclr

Milestone:-

kunalspathak
kunalspathak commented Jun 2, 2023
View reviewed changes
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
jkotas
jkotas reviewed Jun 3, 2023
View reviewed changes
Comment threadsrc/coreclr/inc/corinfo.h Outdated
jkotas
jkotas reviewed Jun 3, 2023
View reviewed changes
Comment threadsrc/coreclr/vm/jitinterface.cpp Outdated
jkotas
jkotas reviewed Jun 3, 2023
View reviewed changes
Comment threadsrc/coreclr/vm/jitinterface.cpp Outdated
@kunalspathakkunalspathak mentioned this pull request Jun 8, 2023
Merged
jkotas
jkotas reviewed Jun 30, 2023
View reviewed changes
Comment threadsrc/coreclr/jit/helperexpansion.cpp
BruceForstall
BruceForstall suggested changes Jul 1, 2023
View reviewed changes
Comment threadsrc/coreclr/tools/superpmi/superpmi-shared/methodcontext.cpp Outdated
Comment threadsrc/coreclr/tools/superpmi/superpmi-shared/methodcontext.cpp Outdated
Comment threadsrc/coreclr/tools/superpmi/superpmi-shared/methodcontext.cpp Outdated
Comment threadsrc/coreclr/tools/superpmi/superpmi-shared/methodcontext.cpp Outdated
Comment threadsrc/coreclr/tools/superpmi/superpmi-shared/methodcontext.cpp Outdated
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
JITDUMP("offsetOfThreadLocalStoragePointer= %u\n", threadStaticBlocksInfo.offsetOfThreadLocalStoragePointer);
JITDUMP("offsetOfMaxThreadStaticBlocks= %u\n", offsetOfMaxThreadStaticBlocksVal);
JITDUMP("offsetOfThreadStaticBlocks= %u\n", offsetOfThreadStaticBlocksVal);
JITDUMP("tlsIndex= %u\n", (ssize_t)threadStaticBlocksInfo.tlsIndex.addr);

@BruceForstallBruceForstallJul 1, 2023

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Dump threadStaticBlocksInfo.tlsIndex.accessType also?

@kunalspathakkunalspathakJul 2, 2023

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you think it will be useful? I don't see we print accessType elsewhere.

@BruceForstallBruceForstallJul 3, 2023

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you think it will be useful? I don't see we print accessType elsewhere.

I don't know. Structurally, it seems odd to omit it. Up to you.

@ghostghost added needs-author-action An issue or pull request that requires more info or actions from the author. and removed needs-author-action An issue or pull request that requires more info or actions from the author. labels Jul 1, 2023
This was referenced Jul 3, 2023
Closed
Closed
BruceForstall
BruceForstall reviewed Jul 3, 2023
View reviewed changes
Comment threadsrc/coreclr/jit/gentree.cpp Outdated
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
Comment threadsrc/coreclr/jit/lsraarmarch.cpp Outdated
Comment threadsrc/coreclr/jit/lsraxarch.cpp Outdated
jakobbotsch
jakobbotsch previously requested changes Jul 3, 2023
View reviewed changes
Comment threadsrc/coreclr/jit/helperexpansion.cpp Outdated
Comment threadsrc/coreclr/jit/lsraxarch.cpp Outdated
@ghostghost added needs-author-action An issue or pull request that requires more info or actions from the author. and removed needs-author-action An issue or pull request that requires more info or actions from the author. labels Jul 3, 2023
@kunalspathak

kunalspathak commented Jul 5, 2023

Copy link
Copy Markdown
ContributorAuthor

Ping @BruceForstall

BruceForstall
BruceForstall approved these changes Jul 5, 2023
View reviewed changes

@BruceForstallBruceForstall left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM w/ one nit

Comment threadsrc/coreclr/jit/lsraarmarch.cpp Outdated
@kunalspathak
kunalspathak merged commit a3981dd into dotnet:mainJul 6, 2023
@kunalspathak
kunalspathak deleted the tls_linux branch July 6, 2023 14:35
clamp03 added a commit to clamp03/runtime that referenced this pull request Jul 10, 2023
@clamp03
[RISC-V] Disable Inline TLS field access
2d957d3
- Disable dotnet#87082 for RISC-V
@clamp03clamp03 mentioned this pull request Jul 10, 2023
Merged
jkotas pushed a commit that referenced this pull request Jul 10, 2023
@clamp03
[RISC-V] Disable Inline TLS field access (#88584)
caff01c
- Disable #87082 for RISC-V
@kunalspathakkunalspathak mentioned this pull request Jul 11, 2023
Closed
@EgorBoEgorBo mentioned this pull request Jul 13, 2023
Closed
@jakobbotschjakobbotsch mentioned this pull request Jul 18, 2023
Closed
@am11am11 mentioned this pull request Jul 26, 2023
Merged
@ghostghost locked as resolved and limited conversation to collaborators Aug 5, 2023
Sign up for freeto subscribe to this conversation on GitHub. Already have an account? Sign in.

Reviewers

@jkotasjkotasjkotas left review comments
@jakobbotschjakobbotschjakobbotsch left review comments
@MichalStrehovskyMichalStrehovskyAwaiting requested review from MichalStrehovskyMichalStrehovsky is a code owner
+1 more reviewer
@BruceForstallBruceForstallBruceForstall approved these changes
Reviewers whose approvals may not affect merge requirements

Assignees

@kunalspathakkunalspathak

Labels

area-CodeGen-coreclrCLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@kunalspathak@jkotas@jakobbotsch@BruceForstall