X-Git-Url: https://notaz.gp2x.de/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=drivers%2Fgp2x%2Fasmutils.s;h=3db4e303e352270eec9b3a052ea44fd8ec3d6d97;hb=b2b95d2e0d1fd5e52d03c2152605b09b024c1d0e;hp=917bb0892f9771a004222b3bc5cff090338f288a;hpb=c0bf6f9f02a2b6afb961a7e9195e2168d7e9cecf;p=fceu.git

diff --git a/drivers/gp2x/asmutils.s b/drivers/gp2x/asmutils.s
index 917bb08..3db4e30 100644
--- a/drivers/gp2x/asmutils.s
+++ b/drivers/gp2x/asmutils.s
@@ -8,6 +8,65 @@ flushcache:
     mov pc, lr
 
 
+.global block_or @ void *src, size_t n, int pat
+
+block_or:
+    stmfd   sp!, {r4-r5}
+    orr     r2, r2, r2, lsl #8
+    orr     r2, r2, r2, lsl #16
+    mov     r1, r1, lsr #4
+block_loop_or:
+    ldmia   r0, {r3-r5,r12}
+    subs    r1, r1, #1
+    orr     r3, r3, r2
+    orr     r4, r4, r2
+    orr     r5, r5, r2
+    orr     r12,r12,r2
+    stmia   r0!, {r3-r5,r12}
+    bne     block_loop_or
+    ldmfd   sp!, {r4-r5}
+    bx      lr
+
+
+.global block_andor @ void *src, size_t n, int andpat, int orpat
+
+block_andor:
+    stmfd   sp!, {r4-r6}
+    orr     r2, r2, r2, lsl #8
+    orr     r2, r2, r2, lsl #16
+    orr     r3, r3, r3, lsl #8
+    orr     r3, r3, r3, lsl #16
+    mov     r1, r1, lsr #4
+block_loop_andor:
+    ldmia   r0, {r4-r6,r12}
+    subs    r1, r1, #1
+    and     r4, r4, r2
+    orr     r4, r4, r3
+    and     r5, r5, r2
+    orr     r5, r5, r3
+    and     r6, r6, r2
+    orr     r6, r6, r3
+    and     r12,r12,r2
+    orr     r12,r12,r3
+    stmia   r0!, {r4-r6,r12}
+    bne     block_loop_andor
+    ldmfd   sp!, {r4-r6}
+    bx      lr
+
+
+.global spend_cycles @ c
+
+spend_cycles:
+    mov     r0, r0, lsr #2  @ 4 cycles/iteration
+    sub     r0, r0, #2      @ entry/exit/init
+.sc_loop:
+    subs    r0, r0, #1
+    bpl     .sc_loop
+
+    bx      lr
+
+
+
 /* buggy and slow, probably because function call overhead
 @ renderer helper, based on bitbank's method
 .global draw8pix @ uint8 *P, uint8 *C, uint8 *PALRAM @ dest, src, pal