Skip to content

GitLab

  • Projects
  • Groups
  • Snippets
  • Help
    • Loading...
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
  • Sign in / Register
GHC
GHC
  • Project overview
    • Project overview
    • Details
    • Activity
    • Releases
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributors
    • Graph
    • Compare
    • Locked Files
  • Issues 4,261
    • Issues 4,261
    • List
    • Boards
    • Labels
    • Service Desk
    • Milestones
    • Iterations
  • Merge Requests 419
    • Merge Requests 419
  • Requirements
    • Requirements
    • List
  • CI / CD
    • CI / CD
    • Pipelines
    • Jobs
    • Schedules
  • Security & Compliance
    • Security & Compliance
    • Dependency List
    • License Compliance
  • Operations
    • Operations
    • Incidents
    • Environments
  • Analytics
    • Analytics
    • CI / CD
    • Code Review
    • Insights
    • Issue
    • Repository
    • Value Stream
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Members
    • Members
  • Collapse sidebar
  • Activity
  • Graph
  • Create a new issue
  • Jobs
  • Commits
  • Issue Boards
  • Glasgow Haskell Compiler
  • GHCGHC
  • Issues
  • #18948

Closed
Open
Opened Nov 13, 2020 by Ben Gamari@bgamari🐢Maintainer

Eliminate eventlog way?

I have recently started to question the value of having the eventlog as a configurable option. It's an incredibly useful tool with little to no cost and the observability that is enables is something that is increasingly expected by users.

In general I wonder whether it wouldn't be best to :

  • drop the _l RTS ways, and
  • enable TRACING in the remaining ways
  • deprecate the -eventlog compiler flag

This would bring a few benefits:

  • users wouldn't need to opt-in to eventlogging
  • binary distributions get a bit smaller
  • we can drop a bit of build system complexity

The only cost is that all Haskell executables get larger by around 400kB:

-rw-r--r-- 1 ben users 6.2M Nov 12 09:55 _build/stage1/rts/build/libHSrts-1.0.a
-rw-r--r-- 1 ben users 6.6M Nov 12 09:55 _build/stage1/rts/build/libHSrts-1.0_l.a

-rwxr-xr-x 1 ben users 3.5M Nov 12 09:55 _build/stage1/rts/build/libHSrts-1.0-ghc9.1.0.20201111.so
-rwxr-xr-x 1 ben users 3.9M Nov 12 09:55 _build/stage1/rts/build/libHSrts-1.0_l-ghc9.1.0.20201111.so

In the age of terabytes of storage, this seems like a reasonable trade-off.

Thoughts?

Edited Nov 13, 2020 by Ben Gamari
Assignee
Assign to
9.2.1
Milestone
9.2.1
Assign milestone
Time tracking
None
Due date
None
Reference: ghc/ghc#18948