[PATCH 0/2] Fix possible buffer overflow in remove_subtree()

These patches are proposed for maint (but also apply cleanly to master).

I presume that this is exploitable via Git commands, though I haven't verified it explicitly [1].

I *think* that the rest of the file is OK. open_output_fd() initially looks suspicious, because it strcpy()s a string onto the end of its path argument. But that is only done when to_tempfile is set, which in turn is handled consistently up the callstack up to the point where it is initially set in checkout_entry() if topath is not NULL. So as long as the caller obeys checkout_entry()'s docstring and passes a long enough buffer for topath, I think everything is OK. In any case, the string appended in open_output_fd() is not under the control of the user, so even if there were a bug in this code path it shouldn't be exploitable.

[1] For example, it is conceivable that there are some checks when
    writing a tree that prevent files with such long names from being
    written by Git.  But even if so, it is clearly a bug that could be
    hit locally on any filesystem where PATH_MAX is not a hard limit.

Michael Haggerty (2):
  checkout_entry(): use the strbuf throughout the function
  entry.c: fix possible buffer overflow in remove_subtree()

 entry.c | 64 ++++++++++++++++++++++++++++++++--------------------------------
 1 file changed, 32 insertions(+), 32 deletions(-)