Bug 197062 – Add option to dump JIT memory

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Tadeu Zagallo]

...

[Tadeu Zagallo]

Created attachment 367762 [details]
Patch

[Tadeu Zagallo]

<rdar://problem/49744332>

[Saam Barati]

Comment on attachment 367762 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=367762&action=review

> Source/JavaScriptCore/assembler/LinkBuffer.cpp:264
> +        dumpJITMemory(outData, outData, m_executableMemory->sizeInBytes());

I think you want m_size here

> Source/JavaScriptCore/jit/ExecutableAllocator.cpp:565
> +            ASSERT(fd != -1);

RELEASE_ASSERT

> Source/JavaScriptCore/jit/ExecutableAllocator.cpp:583
> +        if (UNLIKELY(offset + size > bufferSize))

It’s unlikely this would happen, but you should just make the buffer size the size of the jit memory pool. That way you’re guaranteed there is enough space in it. This code here is wrong if a jitmemcpy copies > 32MB

[Tadeu Zagallo]

Comment on attachment 367762 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=367762&action=review

Thanks for the review!

>> Source/JavaScriptCore/jit/ExecutableAllocator.cpp:583
>> +        if (UNLIKELY(offset + size > bufferSize))
> 
> It’s unlikely this would happen, but you should just make the buffer size the size of the jit memory pool. That way you’re guaranteed there is enough space in it. This code here is wrong if a jitmemcpy copies > 32MB

Yeah, I thought it was unlikely enough that it shouldn't matter, but I'll switch it to use `fixedExecutableMemoryPoolSize` instead.

[Tadeu Zagallo]

Created attachment 367821 [details]
Patch for landing

[WebKit Commit Bot]

Comment on attachment 367821 [details]
Patch for landing

Clearing flags on attachment: 367821

Committed r244470: <https://trac.webkit.org/changeset/244470>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.