[RFC PATCH 0/3] grep: don't add subrepos to in-memory alternates

Make git-grep --recurse-submodules stop adding subrepos to the in-memory alternates list and, instead, pass a reference to the subrepo struct down to the threads.

This series is based on ew/hashmap (it uses the added container_of macro).

There're some possibly needed changes I'm still unsure about. All feedback will be highly appreciated:

- textconv cache is written to the_repository's object database even for
  submodules. Should it perhaps be written to submodules' odb instead?

- Considering the following call chain: grep_source_load_driver() >
  userdiff_find_by_path() > git_check_attr() > collect_some_attrs() >
  prepare_attr_stack() > bootstrap_attr_stack():

  * The last function tries to read the attributes from the
    .gitattributes and .git/info/attributes files of the_repository.
    However, for paths inside the submodule, shouldn't it try to read
    these files from the submodule?

  * This function will also call: read_attr() > read_attr_from_index() >
    read_blob_data_from_index() which might, in turn, call
    read_object_file(). Shouldn't we pass the subrepo to it so that it
    can call repo_read_object_file()? (Again, for paths inside the
    submodule, read_object_file() won't be able to find the object as
    we won't be adding to alternates anymore.)

Matheus Tavares (3):
  diff: use the given repo at diff_populate_filespec()
  object: allow parse_object_or_die() to handle any repo
  grep: don't add submodules to the alternates list

 builtin/grep.c  | 82 +++++++++++++++++++++++++++++++++++--------------
 builtin/prune.c |  4 +--
 bundle.c        | 13 +++++---
 diff.c          |  2 +-
 grep.c          | 26 ++++++++--------
 object.c        |  5 +--
 object.h        |  4 ++-
 pack-bitmap.c   |  5 +--
 reachable.c     |  4 +--
 upload-pack.c   |  2 +-
 10 files changed, 94 insertions(+), 53 deletions(-)