Prefer packed representation for CompiledByteCode
As there are many 'CompiledByteCode' objects alive during a GHCi session, representing its element in a more packed manner improves space behaviour at a minimal cost.
When running GHCi on the agda codebase, we find around 380 live 'CompiledByteCode' objects. Packing their respective 'UnlinkedByteCode' can save quite some pointers.
In experiments on the agda codebase, the number of live bytes was reduced by
around 15MB, which is roughly 2% of the overall live bytes.
These numbers have been obtained using -hT -i1 -l.
Thank you for your contribution to GHC!
Please read the checklist below to make sure your contribution fulfills these expectations. Also please answer the following question in your MR description:
Where is the key part of this patch? That is, what should reviewers look at first?
Please take a few moments to address the following points:
-
if your MR may break existing programs (e.g. touches baseor causes the compiler to reject programs), please describe the expected breakage and add the user-facing label. This will run ghc/head.hackage> to characterise the effect of your change on Hackage. -
ensure that your commits are either individually buildable or squashed -
ensure that your commit messages describe what they do (referring to tickets using #NNNNsyntax when appropriate) -
have added source comments describing your change. For larger changes you likely should add a Note and cross-reference it from the relevant places. -
add a testcase to the testsuite. -
updates the users guide if applicable -
mentions new features in the release notes for the next release
If you have any questions don't hesitate to open your merge request and inquire
in a comment. If your patch isn't quite done yet please do add prefix your MR
title with WIP:.
By default a minimal validation pipeline is run on each merge request, the full-ci label can be applied to perform additional validation checks if your MR affects a more unusual configuration.
Once your change is ready please remove the WIP: tag and wait for review. If
no one has offered a review in a few days then please leave a comment mentioning
@triagers and apply the Blocked on Review label.
Merge request reports
Activity
assigned to @fendor
49 51 -- ----------------------------------------------------------------------------- 50 52 -- Compiled Byte Code 51 53 54 type UnlinkedBCOArr = Array Int UnlinkedBCO 55 52 56 data CompiledByteCode = CompiledByteCode 53 { bc_bcos :: [UnlinkedBCO] -- Bunch of interpretable bindings 57 { bc_bcos :: UnlinkedBCOArr -- Bunch of interpretable bindings 54 58 , bc_itbls :: ItblEnv -- A mapping from DataCons to their itbls 55 59 , bc_ffis :: [FFIInfo] -- ffi blocks we allocated 
64 68 instance Outputable CompiledByteCode where 65 69 ppr CompiledByteCode{..} = ppr bc_bcos 66 70 71 instance Outputable (Array Int UnlinkedBCO) where -
A newtype wrapper is considerable less ergonomic to use as we can't use
Foldableinstances, etc...How about
newtype BCOArr a = BCOArr (Array Int a)Is that better?
Edited by Hannes Siebenhandl changed this line in version 2 of the diff
49 51 -- ----------------------------------------------------------------------------- 50 52 -- Compiled Byte Code 51 53 54 type UnlinkedBCOArr = Array Int UnlinkedBCO 55 52 56 data CompiledByteCode = CompiledByteCode 53 { bc_bcos :: [UnlinkedBCO] -- Bunch of interpretable bindings 57 { bc_bcos :: UnlinkedBCOArr -- Bunch of interpretable bindings changed this line in version 2 of the diff
added 1 commit
- d248a67c - Prefer packed representation for CompiledByteCode
added 1 commit
- c0d319e6 - Prefer packed representation for CompiledByteCode
added 1 commit
- fcd5acb9 - Prefer packed representation for CompiledByteCode
added 1 commit
- 42564352 - Prefer packed representation for CompiledByteCode
-
I think it would be best if we replace
BCOArrwithFlatBagfrom !12170 (closed). 289 289 forceModIface. 290 290 -} 291 291 292 293 {- Note [Storing flat structures in 'ModIface'] changed this line in version 7 of the diff
-
Is this blocked on !12170 (closed) ?
added 203 commits
-
42564352...1324b862 - 202 commits from branch
ghc:master - cf760052 - Prefer packed representation for CompiledByteCode
-
42564352...1324b862 - 202 commits from branch
added 1 commit
- 1b1d37c6 - Prefer packed representation for CompiledByteCode
added 1 commit
- ce39e4d6 - Prefer packed representation for CompiledByteCode
added 1 commit
- 13b0d8ef - Prefer packed representation for CompiledByteCode
assigned to @marge-bot and unassigned @fendor
added 5 commits
-
13b0d8ef...f2cc1107 - 4 commits from branch
ghc:master - 427bc5e2 - Prefer packed representation for CompiledByteCode
-
13b0d8ef...f2cc1107 - 4 commits from branch
added 1 commit
- 0e00847f - Prefer packed representation for CompiledByteCode