From 36ed5d5a78a79fb2cc62bf974919ee071d685f62 Mon Sep 17 00:00:00 2001
From: Moinak Ghosh <moinakg@gmail.com>
Date: Thu, 5 Dec 2013 22:20:01 +0530
Subject: [PATCH] Use adapt2 as default compression in archive mode. Add more
 filter auto-selection by compression level in archive mode. Replace odd
 stride lengths in Delta2 with standard numeric type lengths and improve
 performance.

---
 Makefile.in             |  2 +-
 filters/delta2/delta2.c |  7 ++++---
 filters/dispack/dis.cpp | 19 +++++++++++++++++++
 pcompress.c             | 20 ++++++++++++++++++--
 4 files changed, 42 insertions(+), 6 deletions(-)

diff --git a/Makefile.in b/Makefile.in
index 59e17c8..88eaecd 100644
--- a/Makefile.in
+++ b/Makefile.in
@@ -302,7 +302,7 @@ $(PJPGOBJS): $(PJPGSRCS) $(PJPGHDRS)
 	$(VEC_FLAGS) -DBUILD_LIB $(COMMON_CPPFLAGS_cpp) $(@:.o=.cpp) -o $@
 
 $(DISPACKOBJS): $(DISPACKSRCS) $(DISPACKHDRS)
-	$(COMPILE_cpp) $(COMMON_VEC_FLAGS) @SSE_OPT_FLAGS@ -O2 -fsched-spec-load \
+	$(COMPILE_cpp) $(COMMON_VEC_FLAGS) $(DEBUG_STATS_CPPFLAGS) @SSE_OPT_FLAGS@ -O2 -fsched-spec-load \
 	-Wno-variadic-macros $(VEC_FLAGS) $(COMMON_CPPFLAGS_cpp) $(@:.o=.cpp) -o $@
 
 $(SKEIN_BLOCK_OBJ): $(SKEIN_BLOCK_SRC)
diff --git a/filters/delta2/delta2.c b/filters/delta2/delta2.c
index 1bdcea8..9c9387f 100644
--- a/filters/delta2/delta2.c
+++ b/filters/delta2/delta2.c
@@ -91,10 +91,11 @@
 
 /*
  * Stride values to be checked. As of this implementation strides only
- * upto 8 bytes (uint64_t) are supported.
+ * upto 8 bytes (uint64_t) are supported and common type lengths only
+ * are checked.
  */
-#define	NSTRIDES	4
-static uchar_t strides[NSTRIDES] = {3, 5, 7, 8};
+#define	NSTRIDES	3
+static uchar_t strides[NSTRIDES] = {2, 4, 8};
 
 
 static int delta2_encode_real(uchar_t *src, uint64_t srclen, uchar_t *dst, uint64_t *dstlen,
diff --git a/filters/dispack/dis.cpp b/filters/dispack/dis.cpp
index 3c66268..eb49b09 100644
--- a/filters/dispack/dis.cpp
+++ b/filters/dispack/dis.cpp
@@ -29,6 +29,9 @@
 #include <string.h>
 #include <malloc.h>
 #include <assert.h>
+#include <iostream>
+
+using namespace std;
 
 /* Version history:
  *
@@ -944,10 +947,16 @@ dispack_encode(uchar_t *from, uint64_t fromlen, uchar_t *to, uint64_t *dstlen)
 {
 	uchar_t *pos, *hdr, type, *pos_to, *to_last;
 	uint64_t len;
+#ifdef	DEBUG_STATS
+	double strt, en;
+#endif
 
 	if (fromlen < DISFILTER_BLOCK)
 		return (-1);
 
+#ifdef	DEBUG_STATS
+	strt = get_wtime_millis();
+#endif
 	pos = from;
 	len = fromlen;
 	pos_to = to;
@@ -1009,9 +1018,19 @@ dispack_encode(uchar_t *from, uint64_t fromlen, uchar_t *to, uint64_t *dstlen)
 		len -= sz;
 	}
 	*dstlen = pos_to - to;
+#ifdef	DEBUG_STATS
+	en = get_wtime_millis();
+	cerr << "Dispack: Processed at " << get_mb_s(fromlen, strt, en) << " MB/s" << endl;
+#endif
 	if ((fromlen - *dstlen) < DIS_MIN_REDUCE) {
+#ifdef	DEBUG_STATS
+		cerr << "Dispack: Failed, reduction too less" << endl;
+#endif
 		return (-1);
 	}
+#ifdef	DEBUG_STATS
+	cerr << "Dispack: srclen: " << fromlen << ", dstlen: " << *dstlen << endl;
+#endif
 	return (0);
 }
 
diff --git a/pcompress.c b/pcompress.c
index 724d661..2237fb5 100644
--- a/pcompress.c
+++ b/pcompress.c
@@ -2946,6 +2946,7 @@ init_pc_context(pc_ctx_t *pctx, int argc, char *argv[])
 
 		    case 'a':
 			pctx->archive_mode = 1;
+			pctx->do_compress = 1;
 			break;
 
 		    case 'v':
@@ -2978,6 +2979,19 @@ init_pc_context(pc_ctx_t *pctx, int argc, char *argv[])
 		return (2);
 	}
 
+	if (pctx->archive_mode && pctx->do_uncompress) {
+		log_msg(LOG_ERR, 0, "'-a' flag is only for archive creation.");
+		return (1);
+	}
+
+	/*
+	 * Default compression algorithm during archiving is Adaptive2.
+	 */
+	if (pctx->archive_mode && pctx->algo == NULL) {
+		pctx->algo = "adapt2";
+		init_algo(pctx, pctx->algo, 1);
+	}
+
 	if (pctx->level == -1 && pctx->do_compress) {
 		if (memcmp(pctx->algo, "lz4", 3) == 0) {
 			pctx->level = 1;
@@ -3204,16 +3218,18 @@ init_pc_context(pc_ctx_t *pctx, int argc, char *argv[])
 		}
 
 		/*
-		 * Selectively enable filters while compressing.
+		 * Selectively enable filters while archiving, depending on compression level.
 		 */
 		if (pctx->archive_mode) {
 			struct filter_flags ff;
 
 			ff.enable_packjpg = 0;
-			if (pctx->level > 9) ff.enable_packjpg = 1;
+			if (pctx->level > 10) ff.enable_packjpg = 1;
 			init_filters(&ff);
 			pctx->enable_packjpg = ff.enable_packjpg;
 			if (pctx->level > 8) pctx->dispack_preprocess = 1;
+			if (pctx->level > 4) pctx->enable_delta2_encode = 1;
+			if (pctx->level > 9) pctx->lzp_preprocess = 1;
 		}
 		if (pctx->lzp_preprocess || pctx->enable_delta2_encode || pctx->dispack_preprocess) {
 			pctx->preprocess_mode = 1;