--- a/.hgignore	Fri Sep 16 15:18:28 2011 +0200
+++ b/.hgignore	Mon Sep 19 17:39:53 2011 +0200
@@ -1,4 +1,7 @@
 syntax: glob
 *.o
+*.a
 peck_test_x86
-peck_test_arm
+peck_test_neon
+peck_test_neon2
+peck_test_neon4

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Makefile	Mon Sep 19 17:39:53 2011 +0200
@@ -0,0 +1,23 @@
+CFLAGS += -I.
+
+OBJS = peck_fft.o peck_fftr.o
+
+all: libpeck_fft.a
+
+libpeck_fft.a: $(OBJS)
+	$(AR) ruv $@ $(OBJS)
+
+clean:
+	rm -f *.o *.a peck_test
+
+peck_test: peck_test.o
+	$(CC) -o $@ peck_test.o -L. -lpeck_fft -lm
+
+test: peck_test
+	./peck_test
+
+install: libpeck_fft.a
+	mkdir -p $(DESTDIR)/usr/lib
+	mkdir -p $(DESTDIR)/usr/include
+	cp libpeck_fft.a $(DESTDIR)/usr/lib
+	cp peck_fft.h peck_fftr.h _peck_fft_guts.h $(DESTDIR)/usr/include

--- a/compile.sh	Fri Sep 16 15:18:28 2011 +0200
+++ b/compile.sh	Mon Sep 19 17:39:53 2011 +0200
@@ -1,13 +1,3 @@
-
-/opt/arm-2011.03/bin/arm-none-linux-gnueabi-gcc \
-	-O3 -mcpu=cortex-a8 -mfpu=neon -mfloat-abi=softfp -ffast-math -fomit-frame-pointer \
-    -DUSE_SIMD=SIMD_NEON2 \
-    -I . \
-    -o peck_test_arm \
-    peck_fftr.c peck_fft.c \
-    peck_test.c \
-    -lm 
-
 gcc \
     -O2 -march=native -msse2 -mfpmath=sse -ffast-math -fomit-frame-pointer \
     -DUSE_SIMD=SIMD_SSE2 \
@@ -17,6 +7,30 @@
     peck_test.c \
     -lm 
 
-# time ./peck_test_x86
+/opt/arm-2011.03/bin/arm-none-linux-gnueabi-gcc \
+	-O3 -mcpu=cortex-a8 -mfpu=neon -mfloat-abi=softfp -ffast-math -fomit-frame-pointer \
+    -DUSE_SIMD=SIMD_NEON \
+    -I . \
+    -o peck_test_neon \
+    peck_fftr.c peck_fft.c \
+    peck_test.c \
+    -lm 
 
-scp peck_test_arm root@192.168.233.104:.
+/opt/arm-2011.03/bin/arm-none-linux-gnueabi-gcc \
+	-O3 -mcpu=cortex-a8 -mfpu=neon -mfloat-abi=softfp -ffast-math -fomit-frame-pointer \
+    -DUSE_SIMD=SIMD_NEON2 \
+    -I . \
+    -o peck_test_neon2 \
+    peck_fftr.c peck_fft.c \
+    peck_test.c \
+    -lm 
+
+/opt/arm-2011.03/bin/arm-none-linux-gnueabi-gcc \
+	-O3 -mcpu=cortex-a8 -mfpu=neon -mfloat-abi=softfp -ffast-math -fomit-frame-pointer \
+    -DUSE_SIMD=SIMD_NEON4 \
+    -I . \
+    -o peck_test_neon4 \
+    peck_fftr.c peck_fft.c \
+    peck_test.c \
+    -lm 
+

--- a/peck_fft.c	Fri Sep 16 15:18:28 2011 +0200
+++ b/peck_fft.c	Mon Sep 19 17:39:53 2011 +0200
@@ -24,7 +24,7 @@
     const peck_fft_cfg st,
     int m) {
 
-//printf("kf_bfly2\n");
+// printf("kf_bfly2, %d\n", fstride);
 
     peck_fft_cpx *Fout2;
     peck_fft_cpx *tw1 = st->twiddles;
@@ -52,7 +52,7 @@
     const size_t m2=2*m;
     const size_t m3=3*m;
 
-//printf("kf_bfly4\n");
+// printf("kf_bfly4, %d\n", fstride);
 
     tw3 = tw2 = tw1 = st->twiddles;
 
@@ -242,7 +242,7 @@
     const int m = *factors++; /* stage's FFT length / p */
     const peck_fft_cpx *Fout_end = Fout + p*m;
 
-// printf("kf_work\n");
+// printf("kf_work, %d\n", fstride);
 
     if (m == 1) {
         do {

--- a/peck_test.c	Fri Sep 16 15:18:28 2011 +0200
+++ b/peck_test.c	Mon Sep 19 17:39:53 2011 +0200
@@ -25,7 +25,7 @@
     unsigned int i, j;
     peck_fftr_cfg p, pi;
 
-//    enable_runfast();
+    enable_runfast();
 
     const unsigned int N = 256;