Add CONST_ADDR_OBJ, use it in some places (WIP) #1677
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
fingolfin
requested review from
ChrisJefferson,
markuspf,
rbehrends and
stevelinton
ChrisJefferson
approved these changes
Sep 6, 2017
Codecov Report
@@ Coverage Diff @@
## master #1677 +/- ##
==========================================
+ Coverage 64.35% 64.42% +0.06%
==========================================
Files 1002 1003 +1
Lines 326699 326742 +43
Branches 13218 13197 -21
==========================================
+ Hits 210244 210500 +256
+ Misses 113585 113368 -217
- Partials 2870 2874 +4
|
olexandr-konovalov
added
the
gapdays2017-fall
|
I would suggest maybe cleaning up this patch as is, and then it can be merged. Obviously more changes can follow, but don't want to risk it drifting too far and merging getting hard. |
Also remove BAG_HEADER_CONTENTS helper
fingolfin
force-pushed
the
mh/CONST_ADDR_OBJ
branch
from
4a443c8 to
42a48ea
Compare
|
@ChrisJefferson OK... I wasn't sure if this had any worth unless @rbehrends's plan works out... I rebased this and cleaned it up slightly. |
|
Personally I always like const-correctness, and also really hope we can eliminate ward :) |
|
@ChrisJefferson well, then either merge this, or get somebody else to merge it, or close it -- otherwise it'll just rot again... |
fingolfin
added
the
release notes: not needed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR adds new macros
CONST_ADDR_OBJandCONST_PTR_BAG, which work like their counterparts withoutCONST_, but return const pointers. Then, various places in the code are changed to use them. In some places, I added new setter helpers etc.This work is clearly incomplete, tons of more places could be changed.
As to the why: This week, Reimer discovered that ward generates almost no guards for HPC-GAP in
master, see here: gap-system/ward#46 -- this is of course a serious problem, caused by our recent move from macros to static inline functions.Reimer has come up with a plan to tackle this, by essentially getting rid of
ward. Basically, the idea is to a write guard intoPTR_BAGand a read guard intoCONST_PTR_BAG, and then stop usingward(of course more work likely is necessary, but you get the idea). For this to not be unusable due to high overhead, he came up with some clever tricks involving a C language extension supported by GCC and clang for declaring functions pure... I'll leave it to him to explain the details.But even without this, I think there is some merit in making use of
consthere; it might enable some extra optimization opportunities.If people think this PR is in the right direction, I'll clean up the final commit and put in some more work. Thought to get through the whole kernel, I might need some help. We'll see.