#543 closed defect (fixed)
(declare (lambda-lift)) infinite loop.
| Reported by: | Alan Post | Owned by: | felix winkelmann |
|---|---|---|---|
| Priority: | critical | Milestone: | 4.9.0 |
| Component: | compiler | Version: | 4.6.x |
| Keywords: | Cc: | ||
| Estimated difficulty: |
Description
I have done some refactoring to my package genturfa'i. After completing my refactoring, the lambda-lift option now never terminates while compiling genturfa'i.
To recreate this problem:
svn co -r23157 https://code.call-cc.org/svn/chicken-eggs/release/4/genturfahi/trunk cd trunk chicken-install
The problem is occurring in the function compute-extra-variables in optimizer.scm. I narrowed this down using the verbose option to the compiler invoked by chicken-install, but attempts on my part to provide a simpler test case have failed.
Turning off (declare (lambda-lift)) causes genturfa'i to compile normally. I would expect to be able to compile my module with lambda lifting, rather than having in never terminate.
Change History (7)
comment:1 Changed 13 years ago by
| Owner: | set to felix winkelmann |
|---|---|
| Priority: | major → critical |
| Status: | new → assigned |
comment:2 Changed 13 years ago by
comment:3 follow-up: 4 Changed 13 years ago by
I noted that the tests fail with errors of this type:
Warning: error in group outside of tests Error: (length) bad argument type - not a proper list: a-5067
Is this a known problem, or specific to the experimental chicken version?
comment:4 Changed 13 years ago by
Replying to felix:
I noted that the tests fail with errors of this type:
Warning: error in group outside of tests Error: (length) bad argument type - not a proper list: a-5067Is this a known problem, or specific to the
experimentalchicken version?
I've just pulled and rebuilt the latest experimental branch, and made sure I had a clean checkout of genturfa'i. I don't see this problem, I'm sorry to report.
I can't say what is going on as a result. I did not recompile any of my dependent eggs with the latest version of the code, perhaps that is where this problem is manifesting.
I checked the salmonella report for the master branch, and genturfa'i compiled fine last night. The experimental branch hasn't been running for 10 days, so I couldn't check that.
comment:5 Changed 13 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | assigned → closed |
I must have made a mistake. Well, I'll close this ticket then.
Ok, this is the last time this optimization has caused any trouble. I will remove it. Really, now.