Cranelift: Take user stack maps through lowering and emission#8876
Conversation
Previously, user stack maps were inserted by the frontend and preserved in the mid-end. This commit takes them from the mid-end CLIF into the backend vcode, and then from that vcode into the finalized mach buffer during emission. During lowering, we compile the `UserStackMapEntry`s into packed `UserStackMap`s. This is the appropriate moment in time to do that coalescing, packing, and compiling because the stack map entries are immutable from this point on. Additionally, we include user stack maps in the `Debug` and disassembly implementations for vcode, just after their associated safepoint instructions. This allows us to see the stack maps we are generating when debugging, as well as write filetests that check we are generating the expected stack maps for the correct instructions. Co-Authored-By: Trevor Elliott <telliott@fastly.com>
github-actions
Bot
added
cranelift
cfallin
left a comment
cfallin
left a comment
There was a problem hiding this comment.
I really like this and look forward to seeing the original stackmap support deleted soon! (At that point I suppose we can rename "user stack maps" back to simply "stack maps" again?)
Some thoughts below but overall shape LGTM.
|
Thanks for the review, Chris! |
Yeah, next I will get Wasmtime using the new system, and then I will start removing the old system. Not 100% sure if I will spend the time to rename "user stack maps" to "stack maps", as the name kinda makes sense to me, given that these are really provided by the frontend "user" and the "user" is responsible for them (as opposed to being generated by Cranelift itself). But I don't feel strongly about it. |
2 participants
Previously, user stack maps were inserted by the frontend and preserved in the mid-end. This commit takes them from the mid-end CLIF into the backend vcode, and then from that vcode into the finalized mach buffer during emission.
During lowering, we compile the
UserStackMapEntrys into packedUserStackMaps. This is the appropriate moment in time to do that coalescing, packing, and compiling because the stack map entries are immutable from this point on.Additionally, we include user stack maps in the
Debugand disassembly implementations for vcode, just after their associated safepoint instructions. This allows us to see the stack maps we are generating when debugging, as well as write filetests that check we are generating the expected stack maps for the correct instructions.