| [hatari-devel] DSP Mandelbrot bug |
[ Thread Index | Date Index | More lists.tuxfamily.org/hatari-devel Archives ]
Hello Hatari Community, i am experiencing a hard to find DSP bug here with Previous. Luckily it can be made "visible" using NeXTstep's included Mandelbrot demo. It might also be responsible for some distorted audio in other applications. I appended a screenshot of the mandelbrot application where the effect of the bug is clearly visible. I pointed to one failing pixel. I also appended some debugging output containing the calculation of the failing pixel and one pixel before and one after the failing one. The last file i appended contains an overview about the variables during calculation to get a better overview. Short overview on the calculation: Every pixel is calculated separately. The visible pixel color is derived from the remaining loop count of some calculation. The higher the remaining count, the lower is the output value of the function. The loop exits using a jec instruction (check extension bit, exit if false). For the "good" pixels it exits after the second run, because the upper 9 bits of b are no longer all 0. For the "bad" pixel it does not exit, because these bits are still all 0. The most suspect part of the calculation seems to be mpy +x0,x0,a at p:0097. The value of a after the third call of that instruction seems to not fit into the pattern. Can someone with more DSP experience see the bug? It might or might not be in dsp_mul56. Any help is greatly appreciated! Andreas
p:008f - p:0094: setup x0: 0x2bd70a (+ 0x51ec) 0x2c28f6 (+ 0x51eb) 0x2c7ae1 x1: 0x2bd70a 0x2c28f6 0x2c7ae1 a: 00:2bd70a:000000 00:2c28f6:000000 00:2c7ae1:000000 y0: 0x20f5c3 0x20f5c3 0x20f5c3 y1: 0x20f5c3 0x20f5c3 0x20f5c3 b: 00:20f5c3:000000 00:20f5c3:000000 00:20f5c3:000000 sp: 04 04 04 ssh: 96 96 96 ssl: 8050 8050 8050 la: 00a7 00a7 00a7 lc: 0003 0003 0003 do: p:0096: mpy +y0,y0,b a,x0: x0: ==, 38e55f, ff13b5 x0: ==, 39a7ef, fff376 x0: ==, 3a6b4f, 00d874 b: 00:087cb9:5da512, 00:2fabf8:5b9408, 00:7ffffe:000002 b: 00:087cb9:5da512, 00:30132d:154082, 00:7ffffe:000002 b: 00:087cb9:5da512, 00:307acf:d30702, 00:7ffffe:000002 sr: ==, ==, 8040 sr: ==, ==, 8040 sr: ==, ==, 8041 p:0097: mpy +x0,x0,a: a: 00:0f03e3:fb98c8, 00:194a4e:263282, 00:0001b4:34bbf2 a: 00:0f3c36:3b98c8, 00:19f86d:2f6242, 00:000001:3a74c8 a: 00:0f74f0:a67382, 00:1aa999:8e44c2, 00:00016e:07e920 sr: ==, ==, 8050 sr: ==, ==, 8050 sr: ==, ==, 8051 p:0098: add a,b: b: 00:17809d:593dda, 00:48f646:81c68a, 00:8001b2:34bbf4 b: 00:17b8ef:993dda, 00:4a0b9a:44a2c4, 00:7fffff:3a74ca b: 00:17f1aa:041894, 00:4b2469:614bc4, 00:80016c:07e922 sr: ==, 8040, 8060 sr: ==, 8040, 8040 sr: ==, 8040, 8060 p:0099: jec p:$00a1: true, true, false --> enddo true, true, true true, true, false --> enddo p:00a1: mpy +y0,x0,b: b: 00:0b49ee:e2bd3c, 00:22b8e1:dd4f84 b: 00:0b5f07:2d42c4, 00:2355a0:579062, ff:fff376:001914 b: 00:0b741f:35dcc6, 00:23f3de:e304e2 p:00a2: asl b: b: 00:1693dd:c57a78, 00:4571c3:ba9f08 b: 00:16be0e:5a8588, 00:46ab40:af20c4, ff:ffe6ec:003228 b: 00:16e83e:6bb98c, 00:47e7bd:c609c4 p:00a3: asl b: b: 00:2d27bb:8af4f0, 00:8ae387:753e10 b: 00:2d7c1c:b50b10, 00:8d5681:5e4188, ff:ffcdd8:006450 b: 00:2dd07c:d77318, 00:8fcf7b:8c1388 p:00a4: add y1,b: b: 00:4e1d7e:8af4f0, 00:abd94a:753e10 b: 00:4e71df:b50b10, 00:ae4c44:5e4188, 00:20c39b:006450 b: 00:4ec63f:d77318, 00:b0c53e:8c1388 p:00a5: mac -y0,y0,a b,y0 y0: 4e1d7e, 7fffff y0: 4e71df, 7fffff, 20c39b y0: 4ec63f, 7fffff a: 00:06872a:9df3b6, ff:e99e55:ca9e7a a: 00:06bf7c:ddf3b6, ff:e9e540:1a21c0, ff:800003:3a74c6 a: 00:06f837:48ce70, ff:ea2ec9:bb3dc0 p:00a6: asl a: a: 00:0d0e55:3be76c, ff:d33cab:953cf4 a: 00:0d7ef9:bbe76c, ff:d3ca80:344380, ff:000006:74e98c a: 00:0df06e:919ce0, ff:d45d93:767b80 p:00a7: add x1,a: a: 00:38e55f:3be76c, ff:ff13b5:953cf4 a: 00:39a7ef:bbe76c, ff:fff376:344380, ff:2c28fc:74e98c a: 00:3a6b4f:919ce0, 00:00d874:767b80 loop back --> p:0096 (or continue at p:00a8 if loop counter 0)
Mem: x:0x31cc 0x000002 -> 0x000002 p:008f 44d900 (02 cyc) move x:(r1)+,x0 Reg: x0 $000001 -> $2bd70a Reg: r1 $2a9a -> $2a9b p:0090 46d100 (02 cyc) move x:(r1)-,y0 Reg: y0 $7fffff -> $20f5c3 Reg: r1 $2a9b -> $2a9a p:0091 044911 (04 cyc) lua (r1)+n1,r1 Reg: r1 $2a9a -> $2a9c p:0092 208541 (02 cyc) tfr x0,a x0,x1 Reg: x1 $2b851f -> $2bd70a Reg: a $00:000002:000000 -> $00:2bd70a:000000 p:0093 20c759 (02 cyc) tfr y0,b y0,y1 Reg: b $00:80064e:53e34a -> $00:20f5c3:000000 p:0094 06d500 0000a7 (06 cyc) do r5,p:$00a7 Reg: sp $02 -> $04 Reg: ssh $8f -> $96 Reg: ssl $10 -> $8050 Reg: la $00a9 -> $00a7 Reg: lc $00f3 -> $0003 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: b $00:20f5c3:000000 -> $00:087cb9:5da512 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:2bd70a:000000 -> $00:0f03e3:fb98c8 p:0098 200018 (02 cyc) add a,b Reg: b $00:087cb9:5da512 -> $00:17809d:593dda p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:17809d:593dda -> $00:0b49ee:e2bd3c p:00a2 20003a (02 cyc) asl b Reg: b $00:0b49ee:e2bd3c -> $00:1693dd:c57a78 p:00a3 20003a (02 cyc) asl b Reg: b $00:1693dd:c57a78 -> $00:2d27bb:8af4f0 p:00a4 200078 (02 cyc) add y1,b Reg: b $00:2d27bb:8af4f0 -> $00:4e1d7e:8af4f0 Reg: sr $8050 -> $8040 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $20f5c3 -> $4e1d7e Reg: a $00:0f03e3:fb98c8 -> $00:06872a:9df3b6 Reg: sr $8040 -> $8050 p:00a6 200032 (02 cyc) asl a Reg: a $00:06872a:9df3b6 -> $00:0d0e55:3be76c p:00a7 200060 (02 cyc) add x1,a Reg: a $00:0d0e55:3be76c -> $00:38e55f:3be76c p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: x0 $2bd70a -> $38e55f Reg: b $00:4e1d7e:8af4f0 -> $00:2fabf8:5b9408 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:38e55f:3be76c -> $00:194a4e:263282 p:0098 200018 (02 cyc) add a,b Reg: b $00:2fabf8:5b9408 -> $00:48f646:81c68a Reg: sr $8050 -> $8040 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:48f646:81c68a -> $00:22b8e1:dd4f84 Reg: sr $8040 -> $8050 p:00a2 20003a (02 cyc) asl b Reg: b $00:22b8e1:dd4f84 -> $00:4571c3:ba9f08 Reg: sr $8050 -> $8040 p:00a3 20003a (02 cyc) asl b Reg: b $00:4571c3:ba9f08 -> $00:8ae387:753e10 Reg: sr $8040 -> $8060 p:00a4 200078 (02 cyc) add y1,b Reg: b $00:8ae387:753e10 -> $00:abd94a:753e10 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $4e1d7e -> $7fffff Reg: a $00:194a4e:263282 -> $ff:e99e55:ca9e7a Reg: sr $8060 -> $8058 p:00a6 200032 (02 cyc) asl a Reg: a $ff:e99e55:ca9e7a -> $ff:d33cab:953cf4 Reg: sr $8058 -> $8059 p:00a7 200060 (02 cyc) add x1,a Reg: a $ff:d33cab:953cf4 -> $ff:ff13b5:953cf4 Reg: sr $8059 -> $8058 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: x0 $38e55f -> $ff13b5 Reg: b $00:abd94a:753e10 -> $00:7ffffe:000002 Reg: sr $8058 -> $8040 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $ff:ff13b5:953cf4 -> $00:0001b4:34bbf2 Reg: sr $8040 -> $8050 p:0098 200018 (02 cyc) add a,b Reg: b $00:7ffffe:000002 -> $00:8001b2:34bbf4 Reg: sr $8050 -> $8060 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:009b 22ae00 (02 cyc) move r5,a Reg: a $00:0001b4:34bbf2 -> $00:000003:000000 p:009c 0444bf (02 cyc) movec lc,x0 Reg: x0 $ff13b5 -> $000001 p:009d 200044 (02 cyc) sub x0,a Reg: a $00:000003:000000 -> $00:000002:000000 Reg: sr $8060 -> $8050 p:009e 00008c (02 cyc) enddo Reg: sp $04 -> $02 Reg: ssh $96 -> $8f Reg: ssl $8050 -> $10 Reg: la $00a7 -> $00a9 Reg: lc $0001 -> $00f3 p:009f 0af080 0000a9 (06 cyc) jmp p:$00a9 p:00a9 564e00 (02 cyc) move a,x:(r6)+n6 Reg: r6 $31cd -> $31ce Mem: x:0x31cd 0x000002 -> 0x000002 p:008f 44d900 (02 cyc) move x:(r1)+,x0 Reg: x0 $000001 -> $2c28f6 Reg: r1 $2a9c -> $2a9d p:0090 46d100 (02 cyc) move x:(r1)-,y0 Reg: y0 $7fffff -> $20f5c3 Reg: r1 $2a9d -> $2a9c p:0091 044911 (04 cyc) lua (r1)+n1,r1 Reg: r1 $2a9c -> $2a9e p:0092 208541 (02 cyc) tfr x0,a x0,x1 Reg: x1 $2bd70a -> $2c28f6 Reg: a $00:000002:000000 -> $00:2c28f6:000000 p:0093 20c759 (02 cyc) tfr y0,b y0,y1 Reg: b $00:8001b2:34bbf4 -> $00:20f5c3:000000 p:0094 06d500 0000a7 (06 cyc) do r5,p:$00a7 Reg: sp $02 -> $04 Reg: ssh $8f -> $96 Reg: ssl $10 -> $8050 Reg: la $00a9 -> $00a7 Reg: lc $00f2 -> $0003 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: b $00:20f5c3:000000 -> $00:087cb9:5da512 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:2c28f6:000000 -> $00:0f3c36:3b98c8 p:0098 200018 (02 cyc) add a,b Reg: b $00:087cb9:5da512 -> $00:17b8ef:993dda p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:17b8ef:993dda -> $00:0b5f07:2d42c4 p:00a2 20003a (02 cyc) asl b Reg: b $00:0b5f07:2d42c4 -> $00:16be0e:5a8588 p:00a3 20003a (02 cyc) asl b Reg: b $00:16be0e:5a8588 -> $00:2d7c1c:b50b10 p:00a4 200078 (02 cyc) add y1,b Reg: b $00:2d7c1c:b50b10 -> $00:4e71df:b50b10 Reg: sr $8050 -> $8040 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $20f5c3 -> $4e71df Reg: a $00:0f3c36:3b98c8 -> $00:06bf7c:ddf3b6 Reg: sr $8040 -> $8050 p:00a6 200032 (02 cyc) asl a Reg: a $00:06bf7c:ddf3b6 -> $00:0d7ef9:bbe76c p:00a7 200060 (02 cyc) add x1,a Reg: a $00:0d7ef9:bbe76c -> $00:39a7ef:bbe76c p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: x0 $2c28f6 -> $39a7ef Reg: b $00:4e71df:b50b10 -> $00:30132d:154082 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:39a7ef:bbe76c -> $00:19f86d:2f6242 p:0098 200018 (02 cyc) add a,b Reg: b $00:30132d:154082 -> $00:4a0b9a:44a2c4 Reg: sr $8050 -> $8040 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:4a0b9a:44a2c4 -> $00:2355a0:579062 Reg: sr $8040 -> $8050 p:00a2 20003a (02 cyc) asl b Reg: b $00:2355a0:579062 -> $00:46ab40:af20c4 Reg: sr $8050 -> $8040 p:00a3 20003a (02 cyc) asl b Reg: b $00:46ab40:af20c4 -> $00:8d5681:5e4188 Reg: sr $8040 -> $8060 p:00a4 200078 (02 cyc) add y1,b Reg: b $00:8d5681:5e4188 -> $00:ae4c44:5e4188 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $4e71df -> $7fffff Reg: a $00:19f86d:2f6242 -> $ff:e9e540:1a21c0 Reg: sr $8060 -> $8058 p:00a6 200032 (02 cyc) asl a Reg: a $ff:e9e540:1a21c0 -> $ff:d3ca80:344380 Reg: sr $8058 -> $8059 p:00a7 200060 (02 cyc) add x1,a Reg: a $ff:d3ca80:344380 -> $ff:fff376:344380 Reg: sr $8059 -> $8058 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: x0 $39a7ef -> $fff376 Reg: b $00:ae4c44:5e4188 -> $00:7ffffe:000002 Reg: sr $8058 -> $8040 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $ff:fff376:344380 -> $00:000001:3a74c8 Reg: sr $8040 -> $8050 p:0098 200018 (02 cyc) add a,b Reg: b $00:7ffffe:000002 -> $00:7fffff:3a74ca Reg: sr $8050 -> $8040 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:7fffff:3a74ca -> $ff:fff376:001914 Reg: sr $8040 -> $8058 p:00a2 20003a (02 cyc) asl b Reg: b $ff:fff376:001914 -> $ff:ffe6ec:003228 Reg: sr $8058 -> $8059 p:00a3 20003a (02 cyc) asl b Reg: b $ff:ffe6ec:003228 -> $ff:ffcdd8:006450 p:00a4 200078 (02 cyc) add y1,b Reg: b $ff:ffcdd8:006450 -> $00:20c39b:006450 Reg: sr $8059 -> $8051 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $7fffff -> $20c39b Reg: a $00:000001:3a74c8 -> $ff:800003:3a74c6 Reg: sr $8051 -> $8049 p:00a6 200032 (02 cyc) asl a Reg: a $ff:800003:3a74c6 -> $ff:000006:74e98c Reg: sr $8049 -> $8079 p:00a7 200060 (02 cyc) add x1,a Reg: a $ff:000006:74e98c -> $ff:2c28fc:74e98c Reg: sr $8079 -> $8078 p:00a8 22ae00 (02 cyc) move r5,a Reg: a $ff:2c28fc:74e98c -> $00:000003:000000 p:00a9 564e00 (02 cyc) move a,x:(r6)+n6 Reg: r6 $31ce -> $31cf Mem: x:0x31ce 0x000002 -> 0x000003 p:008f 44d900 (02 cyc) move x:(r1)+,x0 Reg: x0 $fff376 -> $2c7ae1 Reg: r1 $2a9e -> $2a9f p:0090 46d100 (02 cyc) move x:(r1)-,y0 Reg: y0 $20c39b -> $20f5c3 Reg: r1 $2a9f -> $2a9e p:0091 044911 (04 cyc) lua (r1)+n1,r1 Reg: r1 $2a9e -> $2aa0 p:0092 208541 (02 cyc) tfr x0,a x0,x1 Reg: x1 $2c28f6 -> $2c7ae1 Reg: a $00:000003:000000 -> $00:2c7ae1:000000 p:0093 20c759 (02 cyc) tfr y0,b y0,y1 Reg: b $00:20c39b:006450 -> $00:20f5c3:000000 p:0094 06d500 0000a7 (06 cyc) do r5,p:$00a7 Reg: sp $02 -> $04 Reg: ssh $8f -> $96 Reg: ssl $10 -> $8078 Reg: la $00a9 -> $00a7 Reg: lc $00f1 -> $0003 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: b $00:20f5c3:000000 -> $00:087cb9:5da512 Reg: sr $8078 -> $8050 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:2c7ae1:000000 -> $00:0f74f0:a67382 p:0098 200018 (02 cyc) add a,b Reg: b $00:087cb9:5da512 -> $00:17f1aa:041894 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:17f1aa:041894 -> $00:0b741f:35dcc6 p:00a2 20003a (02 cyc) asl b Reg: b $00:0b741f:35dcc6 -> $00:16e83e:6bb98c p:00a3 20003a (02 cyc) asl b Reg: b $00:16e83e:6bb98c -> $00:2dd07c:d77318 p:00a4 200078 (02 cyc) add y1,b Reg: b $00:2dd07c:d77318 -> $00:4ec63f:d77318 Reg: sr $8050 -> $8040 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $20f5c3 -> $4ec63f Reg: a $00:0f74f0:a67382 -> $00:06f837:48ce70 Reg: sr $8040 -> $8050 p:00a6 200032 (02 cyc) asl a Reg: a $00:06f837:48ce70 -> $00:0df06e:919ce0 p:00a7 200060 (02 cyc) add x1,a Reg: a $00:0df06e:919ce0 -> $00:3a6b4f:919ce0 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: x0 $2c7ae1 -> $3a6b4f Reg: b $00:4ec63f:d77318 -> $00:307acf:d30702 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:3a6b4f:919ce0 -> $00:1aa999:8e44c2 p:0098 200018 (02 cyc) add a,b Reg: b $00:307acf:d30702 -> $00:4b2469:614bc4 Reg: sr $8050 -> $8040 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:00a1 2000d8 (02 cyc) mpy +y0,x0,b Reg: b $00:4b2469:614bc4 -> $00:23f3de:e304e2 Reg: sr $8040 -> $8050 p:00a2 20003a (02 cyc) asl b Reg: b $00:23f3de:e304e2 -> $00:47e7bd:c609c4 Reg: sr $8050 -> $8040 p:00a3 20003a (02 cyc) asl b Reg: b $00:47e7bd:c609c4 -> $00:8fcf7b:8c1388 Reg: sr $8040 -> $8060 p:00a4 200078 (02 cyc) add y1,b Reg: b $00:8fcf7b:8c1388 -> $00:b0c53e:8c1388 p:00a5 21e696 (02 cyc) mac -y0,y0,a b,y0 Reg: y0 $4ec63f -> $7fffff Reg: a $00:1aa999:8e44c2 -> $ff:ea2ec9:bb3dc0 Reg: sr $8060 -> $8058 p:00a6 200032 (02 cyc) asl a Reg: a $ff:ea2ec9:bb3dc0 -> $ff:d45d93:767b80 Reg: sr $8058 -> $8059 p:00a7 200060 (02 cyc) add x1,a Reg: a $ff:d45d93:767b80 -> $00:00d874:767b80 Reg: sr $8059 -> $8051 p:0096 21c498 (02 cyc) mpy +y0,y0,b a,x0 Reg: x0 $3a6b4f -> $00d874 Reg: b $00:b0c53e:8c1388 -> $00:7ffffe:000002 Reg: sr $8051 -> $8041 p:0097 200080 (02 cyc) mpy +x0,x0,a Reg: a $00:00d874:767b80 -> $00:00016e:07e920 Reg: sr $8041 -> $8051 p:0098 200018 (02 cyc) add a,b Reg: b $00:7ffffe:000002 -> $00:80016c:07e922 Reg: sr $8051 -> $8060 p:0099 0af0a5 0000a1 (06 cyc) jec p:$00a1 p:009b 22ae00 (02 cyc) move r5,a Reg: a $00:00016e:07e920 -> $00:000003:000000 p:009c 0444bf (02 cyc) movec lc,x0 Reg: x0 $00d874 -> $000001 p:009d 200044 (02 cyc) sub x0,a Reg: a $00:000003:000000 -> $00:000002:000000 Reg: sr $8060 -> $8050 p:009e 00008c (02 cyc) enddo Reg: sp $04 -> $02 Reg: ssh $96 -> $8f Reg: ssl $8078 -> $10 Reg: la $00a7 -> $00a9 Reg: lc $0001 -> $00f1 p:009f 0af080 0000a9 (06 cyc) jmp p:$00a9 p:00a9 564e00 (02 cyc) move a,x:(r6)+n6 Reg: r6 $31cf -> $31d0 Mem: x:0x31cf 0x000002 -> 0x000002
| Mail converted by MHonArc 2.6.19+ | http://listengine.tuxfamily.org/ |