Merge pull request #3226 from libgit2/cmn/racy-diff-again

racy-git, the missing link

4 files changed, 65 insertions, 5 deletions

diff --git a/src/diff.c b/src/diff.c
index d7365ef77..cc93f57cd 100644
--- a/src/diff.c
+++ b/src/diff.c
@@ -816,11 +816,11 @@ static int maybe_modified(
 	} else if (git_oid_iszero(&nitem->id) && new_is_workdir) {
 		bool use_ctime = ((diff->diffcaps & GIT_DIFFCAPS_TRUST_CTIME) != 0);
 		bool use_nanos = ((diff->diffcaps & GIT_DIFFCAPS_TRUST_NANOSECS) != 0);
+		git_index *index;
+		git_iterator_index(&index, info->new_iter);
 
 		status = GIT_DELTA_UNMODIFIED;
 
-		/* TODO: add check against index file st_mtime to avoid racy-git */
-
 		if (S_ISGITLINK(nmode)) {
 			if ((error = maybe_modified_submodule(&status, &noid, diff, info)) < 0)
 				return error;
@@ -839,7 +839,8 @@ static int maybe_modified(
 			 !diff_time_eq(&oitem->ctime, &nitem->ctime, use_nanos)) ||
 			oitem->ino != nitem->ino ||
 			oitem->uid != nitem->uid ||
-			oitem->gid != nitem->gid)
+			oitem->gid != nitem->gid ||
+			(index && nitem->mtime.seconds >= index->stamp.mtime))
 		{
 			status = GIT_DELTA_MODIFIED;
 			modified_uncertain = true;
diff --git a/src/index.c b/src/index.c
index 1fb3c48f3..5ce5522f8 100644
--- a/src/index.c
+++ b/src/index.c
@@ -688,20 +688,59 @@ int git_index__changed_relative_to(
 	return !!git_oid_cmp(&index->checksum, checksum);
 }
 
+static bool is_racy_timestamp(git_time_t stamp, git_index_entry *entry)
+{
+	/* Git special-cases submodules in the check */
+	if (S_ISGITLINK(entry->mode))
+		return false;
+
+	/* If we never read the index, we can't have this race either */
+	if (stamp == 0)
+		return false;
+
+	/* If the timestamp is the same or newer than the index, it's racy */
+	return ((int32_t) stamp) <= entry->mtime.seconds;
+}
+
 /*
  * Force the next diff to take a look at those entries which have the
  * same timestamp as the current index.
  */
-static void truncate_racily_clean(git_index *index)
+static int truncate_racily_clean(git_index *index)
 {
 	size_t i;
+	int error;
 	git_index_entry *entry;
 	git_time_t ts = index->stamp.mtime;
+	git_diff_options diff_opts = GIT_DIFF_OPTIONS_INIT;
+	git_diff *diff;
 
+	/* Nothing to do if there's no repo to talk about */
+	if (!INDEX_OWNER(index))
+		return 0;
+
+	/* If there's no workdir, we can't know where to even check */
+	if (!git_repository_workdir(INDEX_OWNER(index)))
+		return 0;
+
+	diff_opts.flags |= GIT_DIFF_INCLUDE_TYPECHANGE | GIT_DIFF_IGNORE_SUBMODULES | GIT_DIFF_DISABLE_PATHSPEC_MATCH;
 	git_vector_foreach(&index->entries, i, entry) {
-		if (entry->mtime.seconds == ts || ts == 0)
+		if (!is_racy_timestamp(ts, entry))
+			continue;
+
+		diff_opts.pathspec.count = 1;
+		diff_opts.pathspec.strings = (char **) &entry->path;
+
+		if ((error = git_diff_index_to_workdir(&diff, INDEX_OWNER(index), index, &diff_opts)) < 0)
+			return error;
+
+		if (git_diff_num_deltas(diff) > 0)
 			entry->file_size = 0;
+
+		git_diff_free(diff);
 	}
+
+	return 0;
 }
 
 int git_index_write(git_index *index)
diff --git a/src/iterator.c b/src/iterator.c
index 7807a1636..d5f7eec34 100644
--- a/src/iterator.c
+++ b/src/iterator.c
@@ -1762,6 +1762,18 @@ int git_iterator_current_workdir_path(git_buf **path, git_iterator *iter)
 	return 0;
 }
 
+int git_iterator_index(git_index **out, git_iterator *iter)
+{
+	workdir_iterator *wi = (workdir_iterator *)iter;
+
+	if (iter->type != GIT_ITERATOR_TYPE_WORKDIR)
+		*out = NULL;
+
+	*out = wi->index;
+
+	return 0;
+}
+
 int git_iterator_advance_over_with_status(
 	const git_index_entry **entryptr,
 	git_iterator_status_t *status,
diff --git a/src/iterator.h b/src/iterator.h
index db1f325a7..57f82416a 100644
--- a/src/iterator.h
+++ b/src/iterator.h
@@ -11,6 +11,7 @@
 #include "git2/index.h"
 #include "vector.h"
 #include "buffer.h"
+#include "ignore.h"
 
 typedef struct git_iterator git_iterator;
 
@@ -286,4 +287,11 @@ typedef enum {
 extern int git_iterator_advance_over_with_status(
 	const git_index_entry **entry, git_iterator_status_t *status, git_iterator *iter);
 
+/**
+ * Retrieve the index stored in the iterator.
+ *
+ * Only implemented for the workdir iterator
+ */
+extern int git_iterator_index(git_index **out, git_iterator *iter);
+
 #endif