HDB+GHC session for HLS and debug#308
Conversation
|
I should also land my changes that introduce the view as HTML DebugView instance for |
Saizan
force-pushed
the
hdb-ghc-debug-on-abstract
branch
3 times, most recently
from
0b8044b to
e06a2b5
Compare
Saizan
force-pushed
the
hdb-ghc-debug-on-abstract
branch
2 times, most recently
from
5d50081 to
c99b8eb
Compare
alt-romes
left a comment
alt-romes
left a comment
There was a problem hiding this comment.
LGTM, just a few tweaks
| - Setup the final debuggee with a cradle that won't use package dbs setup by HIE. | ||
| - See Limitations | ||
| - See ./scripts/cabal-repl-bios.sh for a script that can be used as a bios cradle for this, uses cabal repl. |
There was a problem hiding this comment.
I don't understand this too well. I thought setting the hie.yaml to debug-with-ghc-bios to be sufficient? does this mean we have to do something else?
There was a problem hiding this comment.
Yes, I was getting segfaults when the debugee's debugee was using a normal cradle and using the above script with a bios cradle helped.
So I suppose you can start without and look into this if you get a segfault while the units are getting setup. At some point we'll actually solve this.
| - See https://gitlab.haskell.org/trac-Saizan/ghc/-/commits/debug-9.14 | ||
| - Gone through early stages of prep and build so `./ghc/_build/stageBoot/bin/` is populated. | ||
| - Installed ghc of same version and flavour as GHC source tree being debugged. | ||
| - Installed hdb with executable-dynamic: True. |
There was a problem hiding this comment.
does this hdb have to be compiled by the same ghc as above? If yes say explicitly
Saizan
force-pushed
the
hdb-ghc-debug-on-abstract
branch
from
533c6f9 to
3a0ea80
Compare
2 participants
See DEBUG.md for required setup and limitations