A small test of mine show a tenfold speed improvemnt looking for a blob
while scanning the history (11K commit, blob three trees down). This was
on a x86 platform. For other platforms Shawn noted that memory mapping
was about 10% slower than normall reading of files.
Signed-off-by: Robin Rosenberg <robin.rosenberg@xxxxxxxxxx>
---
.../src/org/spearce/jgit/lib/PackFile.java | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/org.spearce.jgit/src/org/spearce/jgit/lib/PackFile.java b/org.spearce.jgit/src/org/spearce/jgit/lib/PackFile.java
index ee4b8a1..d33aa97 100644
--- a/org.spearce.jgit/src/org/spearce/jgit/lib/PackFile.java
+++ b/org.spearce.jgit/src/org/spearce/jgit/lib/PackFile.java
@@ -42,7 +42,7 @@ public class PackFile {
repo = parentRepo;
// FIXME window size and mmap type should be configurable
pack = new WindowedFile(repo.getWindowCache(), packFile,
- 64 * 1024 * 1024, false);
+ 64 * 1024 * 1024, true);
try {
readPackHeader();
@@ -53,7 +53,7 @@ public class PackFile {
+ ".idx");
// FIXME window size and mmap type should be configurable
idx = new WindowedFile(repo.getWindowCache(), idxFile,
- 64 * 1024 * 1024, false);
+ 64 * 1024 * 1024, true);
try {
idxHeader = readIndexHeader();
} catch (IOException ioe) {
-
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html