Page MenuHomePhabricator

Sonicadvance1 (Ryan Houdek)
User

Projects

User does not belong to any projects.

User Details

User Since
Dec 31 2016, 11:29 PM (256 w, 3 d)

Recent Activity

Jan 26 2021

Sonicadvance1 added a comment to D94954: Fixes Snapdragon Kryo CPU core detection.

This has mixed signals to me.
Looking at the recent commits, there is a smattering of commits putting this in the commit message or not.
Additionally they already have the differential review in the commit messages.

Jan 26 2021, 3:31 PM · Restricted Project

Jan 20 2021

Sonicadvance1 added a comment to D94954: Fixes Snapdragon Kryo CPU core detection.

I don't have commit rights. So someone is going to need to merge this on my behalf.

Jan 20 2021, 8:28 AM · Restricted Project

Jan 19 2021

Sonicadvance1 added a comment to D94954: Fixes Snapdragon Kryo CPU core detection.

There are some tests in llvm/unittests/Support/Host.cpp. Can you add test for these cpu's and the multiple infos you were seeing? I suppose it now gets the info from the last one?

Jan 19 2021, 2:39 PM · Restricted Project
Sonicadvance1 updated the diff for D94954: Fixes Snapdragon Kryo CPU core detection.

Adds unit test to ensure it pulls out cortex-a77 from a Snapdragon mixed implementer minimum cpuinfo.

Jan 19 2021, 2:37 PM · Restricted Project
Sonicadvance1 updated the diff for D94954: Fixes Snapdragon Kryo CPU core detection.

Whitespace fix on Fujitsu check

Jan 19 2021, 2:15 PM · Restricted Project
Sonicadvance1 updated the diff for D94954: Fixes Snapdragon Kryo CPU core detection.

There was a desync between Implementer and Part scans.
Implementer would scan to the end of /proc/cpuinfo, setting the implementer to 0x41 (ARM) but then the StringSwitch loop would return the first definition it would find. Which was CPU part 0x805, which doesn't exist for implementer 0x41 because that was actually a 0x51 implementer part.
Moves the Part scan to the start alongside the Implementer scan so these two values stay in sync across platforms that mix implementers.

Jan 19 2021, 2:11 PM · Restricted Project
Sonicadvance1 added a comment to D94954: Fixes Snapdragon Kryo CPU core detection.

This doesn't fully fix the issue on my end.
I'll need to try again, compiling a simple atomic op with -mcpu=native on my Snapdragon 865 board still did loadstore exclusive instead of CAS. while -mcpu=cortex-a77 does.

Jan 19 2021, 5:49 AM · Restricted Project
Sonicadvance1 requested review of D94954: Fixes Snapdragon Kryo CPU core detection.
Jan 19 2021, 1:39 AM · Restricted Project

Oct 29 2018

Sonicadvance1 added a comment to D52100: [tblgen] Allow FixedLenDecoderEmitter to use APInt-like objects as InsnType.

I'm using APInt directly yes.
Which is then handling a fixed length size of 128bit instructions

Oct 29 2018, 5:05 PM

Oct 26 2018

Sonicadvance1 added a comment to D52100: [tblgen] Allow FixedLenDecoderEmitter to use APInt-like objects as InsnType.

This commit breaks my 128bit instruction target.
I'm currently using a workaround locally for enabling APInt based encodings when your instruction encoding is >64bit.
The main issue with this approach is that getBinaryCodeForInstr completely falls apart.

Could you elaborate on this? getBinaryCodeForInstr() is for encoding instructions whereas FixedLenDecoderEmitter generates a decoder so I'm surprised that you're seeing a connection between getBinaryCodeForInstr() and this patch. Our out-of-tree target didn't use the FixedLenDecoderEmitter until after this patch made it possible but we've been using getBinaryCodeForInstr() for a very long time so I think there's unlikely to be a connection between the two. Additionally, not affecting encoding was a requirement for us as we're also working around the lack of >64-bit support in getBinaryCodeForInstr()

Oct 26 2018, 4:10 PM
Sonicadvance1 added a comment to D52100: [tblgen] Allow FixedLenDecoderEmitter to use APInt-like objects as InsnType.

This commit breaks my 128bit instruction target.
I'm currently using a workaround locally for enabling APInt based encodings when your instruction encoding is >64bit.
The main issue with this approach is that getBinaryCodeForInstr completely falls apart.

Oct 26 2018, 1:40 PM

Mar 17 2018

Sonicadvance1 updated the diff for D44374: [GlobalISel] Add support for lowering vector operations to scalar..

Addressed the comments about readability and added a unit test for the decomposing.
Had to add a specialization that uncovered an issue where vectors with only one element aren't allowed to be generated.

Mar 17 2018, 4:13 AM

Mar 12 2018

Sonicadvance1 added a comment to D44374: [GlobalISel] Add support for lowering vector operations to scalar..

As far as I'm aware it is impossible to write unit tests for this unless a backend uses it?
I've attempted adding this to AArch64's GlobalISel implementation and hit other deficiencies in the backend that are unrelated but this code uncovered.

Mar 12 2018, 7:55 AM
Sonicadvance1 updated the diff for D44374: [GlobalISel] Add support for lowering vector operations to scalar..

Updated the diff with as much context as possible

Mar 12 2018, 7:52 AM

Mar 11 2018

Sonicadvance1 added reviewers for D44374: [GlobalISel] Add support for lowering vector operations to scalar.: dsanders, qcolombet, kristof.beyls, t.p.northover.
Mar 11 2018, 6:19 PM
Sonicadvance1 created D44374: [GlobalISel] Add support for lowering vector operations to scalar..
Mar 11 2018, 6:15 PM