GitLab

Update to match the rest of the docs.

Document -c<n> and bring up to date the documentation of -M<size.

document -ignore-dot-ghci and -read-dot-ghci

Document -package-conf and ghc-pkg --config-file

Allow multiple -f/--config-file options, taking the last one.  This
allows ghc-pkg to be used to modify a user-specified package.conf file.

Update to new syntax and filename-munging method

A bunch of changes (been waiting for the link to cvs.haskell.org to
come back):

- Two new flags -ignore-dot-ghci and -read-dot-ghci control the
  reading (or not) of ./.ghci and $HOME/.ghci.  This will be useful
  for automatic testing of GHCi.

- A new option -package-conf <file> allows reading an additional
  package.conf file, which can be used to keep a per-user set of
  packages.

- GHCi now fails gracefully on startup if linking the libraries
  specified on the command-line fails.

The compilation manager now continues gracefully (by unloading all the
modules) if the link step fails.

resolveObjs now returns a Bool, indicating success

- Convert some fatal errors into non-fatal returns
- Don't include cryptic function names in non-internal errors
- Improve several error messages

The '--' argument disables any further +RTS ... -RTS processing on the
command line.

schemeR_wrk: look through tyapps when spotting top-level nullary
constructor uses.  Fixes bug reported by Mark Tehver:
<<loop>> in GHC 5.01 "Zarjaz"

Don't use packString for GHC > 5.00

Update the hc-build script to the recipe I'm using for the FreeBSD
package build.  The main improvement is that after running this script
the resulting tree should be ready for a 'make install' from the top
level.  This is highly fragile and relies on touching various files to
avoid having to recompile anything that will fail :-)

The FreeBSD package that this derives from is well tested, but I
haven't tested the script itself yet.

Added SysTools.Option, which lets you identify what are
file-path like options to SysTool.run<Foo>.

Using this, we can now precisely control when to transform
filepaths into a host-compatible format (i.e., we can DOSify
just the right bits under Win32).

Fix bug with +RTS -P

Test for identifiers containing non-ASCII characters.

Remove UnsafeCString (normal CString is fast enough now).

Make some foreign imports "unsafe".

Fix another place where the maxHeapSize is set to the
heapSizeSuggestion, now that maxHeapSize == 0 indicates "unbounded".

Fix an assertion: "large" objects may not necessarily start on a block
boundary, but the only ones that are allowed not to are ByteArrays.

Fix Solaris borkage.  Sigh.

Claim that MUT_CONS closures have 1 pointer and 1 non-pointer field,
so that the default threading code in GCCompact works for them (it
doesn't thread the mut_link field).  This also matches MUT_VARs, so
it's more consistent.

Thread the mut_once_list properly, once, before we start compacting.
Previously we tried to do it during the first threading phase, but we
missing some pointers (namely IND_STATIC objects which are only
reachable via the mut_once list).

added pointer to ghc/rts/Adjustor.c, which has some good comments.

add missing symbol to RTS_SYMBOLS

Added SimonM's explanation about f.e.d. stubs.

Remember to initialise stp->bitmap.  Avoids a potential segfault with +RTS -c.

Initialise n_large_blocks and n_scavenged_large_blocks.

Oops, only ASCII alphanumeric characters are allowed to be used
unencoded in C identifiers, but Char.isAlphaNum includes ISO-8851
alphanumeric characters.

Disable debugging machinery which skeaked in in the last commit.

C-side FFI support for Byte/Ptr arrays.

Change "dunno, mate" into something a bit more helpful.

Add support for passing ptr/byte arrays to C.

Fix a bug in gcStablePtrTable().

Disable use of finalisers attached to UnlinkedBCOs, since finalisers
attached to non-atomic objects may run too early :-(

Attach finaliser for malloc'd blocks to the UnlinkedBCOs, not to
linked really-really-really BCOs.  This is because an unlinked BCO
may be copied many times to generated LinkedBCOs before it dies.
Attaching finalisers to linked BCOs could mean multiple free()s on
the same address.

Define allocaBytes in terms of the new pinned ByteArray primitives.
This version is over 6 times faster than the old version using
malloc/free.

update expected output

Remove parse error.  This test actually doesn't work!