Commit 60534e57fae476c35c8d8d20f3e5f5afa690be51

Zefir Kurtisi 2013-12-24T00:33:42

Initial import of Bitmine.ch A1 SPI driver This initial driver was designed around an FPGA based emulator board which provides 2 chained chips with 4 hashing engines each, running at 10% of the A1's nominal speed. Driver has been tested on RasPi. Signed-off-by: Zefir Kurtisi <zefir.kurtisi@gmail.com>

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
diff --git a/ASIC-README b/ASIC-README
index e41a983..75b2030 100644
--- a/ASIC-README
+++ b/ASIC-README
@@ -92,6 +92,11 @@ Currently the U1 USB sticks are supported and come up as ANU devices. They
 are also set up as per the USB ASICs below. They need no options to work well
 but will accept all the icarus options.
 
+
+BITMINE A1 devices
+
+Bitmine A1 devices need the --enable-bitmine_A1 compile option set.
+
 ---
 GETTING STARTED WITH USB ASICS
 
@@ -467,6 +472,19 @@ the temp-overheat feature.
 Other undocumented hashfast command line options are for development purposes
 only at this stage and serve no useful purpose to end users.
 
+
+Bitmine A1 Devices
+
+--bitmine-a1-options <ref_clk>:<sys_clk>:<spi_clk>:<max_chip>
+ref_clk:  reference clock in kHz                      (default: 16000)
+sys_clk:  target system clock in kHz to be set in PLL (default: 250000)
+spi_clk:  SPI clock in kHz                            (default: 800)
+max_chip: [debug/testing] limit chip chain
+
+Set 0 for fields you want to keep untouched to default, e.g.
+--bitmine-a1-options 0:0:400
+to only set SPI clock to 400kHz
+
 ---
 
 This code is provided entirely free of charge by the programmer in his spare
diff --git a/Makefile.am b/Makefile.am
index 265b151..3ed1b39 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -82,6 +82,11 @@ if HAS_BITFURY
 cgminer_SOURCES += driver-bitfury.c driver-bitfury.h
 endif
 
+if HAS_BITMINE_A1
+cgminer_SOURCES += driver-SPI-bitmine-A1.c
+cgminer_SOURCES += spi-context.c spi-context.h
+endif
+
 if HAS_DRILLBIT
 cgminer_SOURCES += driver-drillbit.c driver-drillbit.h
 endif
diff --git a/api.c b/api.c
index 4e0247b..697f4dd 100644
--- a/api.c
+++ b/api.c
@@ -26,7 +26,7 @@
 #include "util.h"
 #include "klist.h"
 
-#if defined(USE_BFLSC) || defined(USE_AVALON) || defined(USE_HASHFAST) || defined(USE_BITFURY) || defined(USE_KLONDIKE) || defined(USE_KNC) || defined(USE_BAB) || defined(USE_DRILLBIT) || defined(USE_MINION) || defined(USE_COINTERRA)
+#if defined(USE_BFLSC) || defined(USE_AVALON) || defined(USE_HASHFAST) || defined(USE_BITFURY) || defined(USE_KLONDIKE) || defined(USE_KNC) || defined(USE_BAB) || defined(USE_DRILLBIT) || defined(USE_MINION) || defined(USE_COINTERRA) || defined(USE_BITMINE_A1)
 #define HAVE_AN_ASIC 1
 #endif
 
@@ -171,6 +171,9 @@ static const char *DEVICECODE = ""
 #ifdef USE_HASHFAST
 			"HFA "
 #endif
+#ifdef USE_BITMINE_A1
+			"BA1 "
+#endif
 #ifdef USE_ICARUS
 			"ICA "
 #endif
diff --git a/cgminer.c b/cgminer.c
index 507012c..453a4d9 100644
--- a/cgminer.c
+++ b/cgminer.c
@@ -202,6 +202,9 @@ char *opt_klondike_options = NULL;
 char *opt_drillbit_options = NULL;
 #endif
 char *opt_bab_options = NULL;
+#ifdef USE_BITMINE_A1
+char *opt_bitmine_a1_options = NULL;
+#endif
 #ifdef USE_USBUTILS
 char *opt_usb_select = NULL;
 int opt_usbdump = -1;
@@ -1128,7 +1131,14 @@ static char *set_bab_options(const char *arg)
 	return NULL;
 }
 #endif
+#ifdef USE_BITMINE_A1
+static char *set_bitmine_a1_options(const char *arg)
+{
+	opt_set_charp(arg, &opt_bitmine_a1_options);
 
+	return NULL;
+}
+#endif
 #ifdef USE_USBUTILS
 static char *set_usb_select(const char *arg)
 {
@@ -1331,6 +1341,11 @@ static struct opt_table opt_config_table[] = {
 		     set_klondike_options, NULL, NULL,
 		     "Set klondike options clock:temptarget"),
 #endif
+#ifdef USE_BITMINE_A1
+	OPT_WITH_ARG("--bitmine-a1-options",
+		     set_bitmine_a1_options, NULL, NULL,
+		     "Bitmine A1 options ref_clk_khz:sys_clk_khz:spi_clk_khz:override_chip_num"),
+#endif
 	OPT_WITHOUT_ARG("--load-balance",
 		     set_loadbalance, &pool_strategy,
 		     "Change multipool strategy from failover to quota based balance"),
@@ -1647,6 +1662,9 @@ static char *opt_verusage_and_exit(const char *extra)
 #ifdef USE_MODMINER
 		"modminer "
 #endif
+#ifdef USE_BITMINE_A1
+		"Bitmine.A1 "
+#endif
 		"mining support.\n"
 		, packagename);
 	printf("%s", opt_usage(opt_argv0, extra));
@@ -4674,6 +4692,10 @@ void write_config(FILE *fcfg)
 #endif
 	if (opt_bab_options)
 		fprintf(fcfg, ",\n\"bab-options\" : \"%s\"", json_escape(opt_bab_options));
+#ifdef USE_BITMINE_A1
+	if (opt_bitmine_a1_options)
+		fprintf(fcfg, ",\n\"bitmine-a1-options\" : \"%s\"", json_escape(opt_bitmine_a1_options));
+#endif
 #ifdef USE_USBUTILS
 	if (opt_usb_select)
 		fprintf(fcfg, ",\n\"usb\" : \"%s\"", json_escape(opt_usb_select));
@@ -7517,7 +7539,7 @@ static void *watchpool_thread(void __maybe_unused *userdata)
 		}
 
 		cgsleep_ms(30000);
-			
+
 	}
 	return NULL;
 }
@@ -8757,7 +8779,7 @@ begin_bench:
 
 		cgpu->rolling = cgpu->total_mhashes = 0;
 	}
-	
+
 	cgtime(&total_tv_start);
 	cgtime(&total_tv_end);
 	get_datestamp(datestamp, sizeof(datestamp), &total_tv_start);
diff --git a/configure.ac b/configure.ac
index a32a066..24d412a 100644
--- a/configure.ac
+++ b/configure.ac
@@ -192,6 +192,17 @@ if test "x$cointerra" = xyes; then
 fi
 AM_CONDITIONAL([HAS_COINTERRA], [test x$cointerra = xyes])
 
+bitmine_A1="no"
+
+AC_ARG_ENABLE([bitmine_A1],
+	[AC_HELP_STRING([--enable-bitmine_A1],[Compile support for Bitmine.ch A1 ASICs (default disabled)])],
+	[bitmine_A1=$enableval]
+	)
+if test "x$bitmine_A1" = xyes; then
+	AC_DEFINE([USE_BITMINE_A1], [1], [Defined to 1 if Bitmine A1 support is wanted])
+fi
+AM_CONDITIONAL([HAS_BITMINE_A1], [test x$bitmine_A1 = xyes])
+
 drillbit="no"
 
 AC_ARG_ENABLE([drillbit],
@@ -203,6 +214,7 @@ if test "x$drillbit" = xyes; then
 fi
 AM_CONDITIONAL([HAS_DRILLBIT], [test x$drillbit = xyes])
 
+
 hashfast="no"
 
 AC_ARG_ENABLE([hashfast],
@@ -513,6 +525,12 @@ if test "x$cointerra" = xyes; then
 	echo "  Cointerra.ASICs......: Enabled"
 else
 	echo "  Cointerra.ASICs......: Disabled"
+
+if test "x$bitmine_A1" = xyes; then
+	echo "  Bitmine-A1.ASICs........: Enabled"
+else
+	echo "  Bitmine-A1.ASICs........: Disabled"
+
 fi
 
 if test "x$drillbit" = xyes; then
@@ -551,7 +569,7 @@ else
 	echo "  ModMiner.FPGAs.......: Disabled"
 fi
 
-if test "x$avalon$bab$bflsc$bitforce$bitfury$hashfast$icarus$klondike$knc$modminer$drillbit$minion$cointerra" = xnonononononononononononono; then
+if test "x$avalon$bab$bflsc$bitforce$bitfury$hashfast$icarus$klondike$knc$modminer$drillbit$minion$cointerra$bitmine_A1" = xnononononononononononononono; then
 	AC_MSG_ERROR([No mining configured in])
 fi
 
diff --git a/driver-SPI-bitmine-A1.c b/driver-SPI-bitmine-A1.c
new file mode 100644
index 0000000..912eebb
--- /dev/null
+++ b/driver-SPI-bitmine-A1.c
@@ -0,0 +1,948 @@
+/*
+ * cgminer SPI driver for Bitmine.ch A1 devices
+ *
+ * Copyright 2013, 2014 Zefir Kurtisi <zefir.kurtisi@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 3 of the License, or (at your option)
+ * any later version.  See COPYING for more details.
+ */
+
+#include <stdlib.h>
+#include <assert.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <unistd.h>
+#include <stdbool.h>
+
+#include "spi-context.h"
+#include "logging.h"
+#include "miner.h"
+#include "util.h"
+
+/********** work queue */
+struct work_ent {
+	struct work *work;
+	struct list_head head;
+};
+
+struct work_queue {
+	int num_elems;
+	struct list_head head;
+};
+
+static bool wq_enqueue(struct work_queue *wq, struct work *work)
+{
+	if (work == NULL)
+		return false;
+	struct work_ent *we = malloc(sizeof(*we));
+	assert(we != NULL);
+
+	we->work = work;
+	INIT_LIST_HEAD(&we->head);
+	list_add_tail(&we->head, &wq->head);
+	wq->num_elems++;
+	return true;
+}
+
+static struct work *wq_dequeue(struct work_queue *wq)
+{
+	if (wq == NULL)
+		return NULL;
+	if (wq->num_elems == 0)
+		return NULL;
+	struct work_ent *we;
+	we = list_entry(wq->head.next, struct work_ent, head);
+	struct work *work = we->work;
+
+	list_del(&we->head);
+	free(we);
+	wq->num_elems--;
+	return work;
+}
+
+static FILE *log_file = NULL;
+#if 1
+#undef applog
+#define applog(X,...) \
+	do { \
+		if (log_file == NULL) { \
+			log_file = fopen("/run/A1.log", "w");\
+			if (log_file==NULL) continue;\
+		}\
+		fprintf(log_file, __VA_ARGS__);\
+		fprintf(log_file, "\n");\
+	} while(0)
+#endif
+
+/********** chip and chain context structures */
+/*
+ * if not cooled sufficiently, communication fails and chip is temporary
+ * disabled. we let it inactive for 10 seconds to cool down
+ *
+ * TODO: to be removed after bring up / test phase
+ */
+#define COOLDOWN_MS (10 * 1000)
+
+/* the WRITE_JOB command is the largest (2 bytes command, 56 bytes payload) */
+#define WRITE_JOB_LENGTH	58
+#define MAX_CHAIN_LENGTH	255
+/*
+ * For commands to traverse the chain, we need to issue dummy writes to
+ * keep SPI clock running. To reach the last chip in the chain, we need to
+ * write the command, followed by chain-length words to pass it through the
+ * chain and another chain-length words to get the ACK back to host
+ */
+#define MAX_CMD_LENGTH		(WRITE_JOB_LENGTH + MAX_CHAIN_LENGTH * 2 * 2)
+
+struct A1_chip {
+	int num_cores;
+	int last_queued_id;
+	struct work *work[4];
+	/* stats */
+	int hw_errors;
+	int stales;
+	int nonces_found;
+	int nonce_ranges_done;
+
+	/* systime in ms when chip was disabled */
+	int cooldown_begin;
+};
+
+struct A1_chain {
+	struct cgpu_info *cgpu;
+	int num_chips;
+	int num_cores;
+	int num_active_chips;
+	uint8_t spi_tx[MAX_CMD_LENGTH];
+	uint8_t spi_rx[MAX_CMD_LENGTH];
+	struct spi_ctx *spi_ctx;
+	struct A1_chip *chips;
+	pthread_mutex_t lock;
+
+	struct work_queue active_wq;
+};
+
+enum A1_command {
+	A1_BIST_START		= 0x01,
+	A1_BIST_FIX		= 0x03,
+	A1_RESET		= 0x04,
+	A1_WRITE_JOB		= 0x07,
+	A1_READ_RESULT		= 0x08,
+	A1_WRITE_REG		= 0x09,
+	A1_READ_REG		= 0x0a,
+	A1_READ_REG_RESP	= 0x1a,
+};
+
+/********** config paramters */
+struct A1_config_options {
+	int ref_clk_khz;
+	int sys_clk_khz;
+	int spi_clk_khz;
+	/* limit chip chain to this number of chips (testing only) */
+	int override_chip_num;
+};
+
+/*
+ * for now, we have one global config, defaulting values:
+ * - ref_clk 16MHz / sys_clk 250MHz
+ * - 800 kHz SPI clock
+ */
+static struct A1_config_options config_options = {
+	.ref_clk_khz = 16000, .sys_clk_khz = 250000, .spi_clk_khz = 800,
+};
+
+/* override values with --bitmine-a1-options ref:sys:spi: - use 0 for default */
+static struct A1_config_options *parsed_config_options;
+
+/********** temporary helper for hexdumping SPI traffic */
+#define DEBUG_HEXDUMP 1
+static void hexdump(char *prefix, uint8_t *buff, int len)
+{
+#if DEBUG_HEXDUMP
+	static char line[256];
+	char *pos = line;
+	int i;
+	if (len < 1)
+		return;
+
+	pos += sprintf(pos, "%s: %d bytes:", prefix, len);
+	for (i = 0; i < len; i++) {
+		if (i > 0 && (i % 32) == 0) {
+			applog(LOG_DEBUG, "%s", line);
+			pos = line;
+			pos += sprintf(pos, "\t");
+		}
+		pos += sprintf(pos, "%.2X ", buff[i]);
+	}
+	applog(LOG_DEBUG, "%s", line);
+#endif
+}
+
+/********** upper layer SPI functions */
+static bool spi_send_command(struct A1_chain *a1, uint8_t cmd, uint8_t addr,
+			     uint8_t *buff, int len)
+{
+	memset(a1->spi_tx + 2, 0, len);
+	a1->spi_tx[0] = cmd;
+	a1->spi_tx[1] = addr;
+	if (len > 0 && buff != NULL)
+		memcpy(a1->spi_tx + 2, buff, len);
+	int tx_len = (2 + len + 1) & ~1;
+	applog(LOG_DEBUG, "Processing command 0x%02x%02x\n", cmd, addr);
+	bool retval = spi_transfer(a1->spi_ctx, a1->spi_tx, a1->spi_rx, tx_len);
+	hexdump("TX", a1->spi_tx, tx_len);
+	hexdump("RX", a1->spi_rx, tx_len);
+	return retval;
+}
+
+static bool spi_poll_result(struct A1_chain *a1, uint8_t cmd,
+			    uint8_t chip_id, int len)
+{
+	int i;
+	int max_poll_words = a1->num_chips * 2 + 1;
+	/* at startup, we don't know the chain-length */
+	if (a1->num_chips == 0)
+		max_poll_words += MAX_CHAIN_LENGTH * 2;
+	for(i = 0; i < max_poll_words; i++) {
+		bool s = spi_transfer(a1->spi_ctx, NULL, a1->spi_rx, 2);
+		hexdump("RX", a1->spi_rx, 2);
+		if (!s)
+			return false;
+		if (a1->spi_rx[0] == cmd && a1->spi_rx[1] == chip_id) {
+			applog(LOG_DEBUG, "Cmd 0x%02x ACK'd", cmd);
+			if (len == 0)
+				return true;
+			s = spi_transfer(a1->spi_ctx, NULL,
+					 a1->spi_rx + 2, len);
+			hexdump("RX", a1->spi_rx + 2, len);
+			if (!s)
+				return false;
+			hexdump("poll_result", a1->spi_rx, len + 2);
+			return true;
+		}
+	}
+	applog(LOG_WARNING, "Failure: missing ACK for cmd 0x%02x", cmd);
+	return false;
+}
+
+/********** A1 SPI commands */
+static bool cmd_BIST_START(struct A1_chain *a1)
+{
+	return	spi_send_command(a1, A1_BIST_START, 0x00, NULL, 0) &&
+		spi_poll_result(a1, A1_BIST_START, 0x00, 2);
+}
+
+static bool cmd_BIST_FIX_BCAST(struct A1_chain *a1)
+{
+	return	spi_send_command(a1, A1_BIST_FIX, 0x00, NULL, 0) &&
+		spi_poll_result(a1, A1_BIST_FIX, 0x00, 0);
+}
+
+static bool cmd_RESET_BCAST(struct A1_chain *a1)
+{
+	return	spi_send_command(a1, A1_RESET, 0x00, NULL, 0) &&
+		spi_poll_result(a1, A1_RESET, 0x00, 0);
+}
+
+static bool cmd_WRITE_REG_BCAST(struct A1_chain *a1, uint8_t *reg)
+{
+	return	spi_send_command(a1, A1_WRITE_REG, 0, reg, 6) &&
+		spi_poll_result(a1, A1_WRITE_REG, 0, 6);
+}
+
+static bool cmd_WRITE_REG(struct A1_chain *a1, uint8_t chip, uint8_t *reg)
+{
+	/* ensure we push the SPI command to the last chip in chain */
+	int tx_length = 8 + a1->num_chips * 4;
+	memcpy(a1->spi_tx, reg, 8);
+	memset(a1->spi_tx + 8, 0, tx_length - 8);
+
+	return spi_transfer(a1->spi_ctx, a1->spi_tx, a1->spi_rx, tx_length);
+}
+
+static bool cmd_READ_REG(struct A1_chain *a1, uint8_t chip)
+{
+	return	spi_send_command(a1, A1_READ_REG, chip, NULL, 0) &&
+		spi_poll_result(a1, A1_READ_REG_RESP, chip, 6);
+}
+
+static bool cmd_WRITE_JOB(struct A1_chain *a1, uint8_t chip_id, uint8_t *job)
+{
+	/* ensure we push the SPI command to the last chip in chain */
+	int tx_length = WRITE_JOB_LENGTH + a1->num_chips * 4;
+	memcpy(a1->spi_tx, job, WRITE_JOB_LENGTH);
+	memset(a1->spi_tx + WRITE_JOB_LENGTH, 0, tx_length - WRITE_JOB_LENGTH);
+
+	return spi_transfer(a1->spi_ctx, a1->spi_tx, a1->spi_rx, tx_length);
+}
+
+/********** A1 low level functions */
+static void A1_hw_reset(void)
+{
+	/*
+	 * TODO: issue cold reset
+	 *
+	 * NOTE: suggested sequence
+	 * a) reset the RSTN pin for at least 1s
+	 * b) release the RSTN pin
+	 * c) wait at least 1s before sending the first CMD
+	 */
+
+#ifdef NOTYET
+	a1_nreset_low();
+	cgsleep_ms(1000);
+	a1_nreset_hi();
+	cgsleep_ms(1000);
+#endif
+}
+
+static bool is_busy(struct A1_chain *a1, uint8_t chip)
+{
+	if (!cmd_READ_REG(a1, chip))
+		return false;
+	return (a1->spi_rx[5] & 0x01) == 0x01;
+}
+
+#define MAX_PLL_WAIT_CYCLES 25
+#define PLL_CYCLE_WAIT_TIME 40
+static bool check_chip_pll_lock(struct A1_chain *a1, int chip_id, uint8_t *wr)
+{
+	int n;
+	for (n = 0; n < MAX_PLL_WAIT_CYCLES; n++) {
+		/* check for PLL lock status */
+		if (cmd_READ_REG(a1, chip_id) && (a1->spi_rx[4] & 1) == 1)
+			/* double check that we read back what we set before */
+			return wr[0] == a1->spi_rx[2] && wr[1] == a1->spi_rx[3];
+
+		cgsleep_ms(40);
+	}
+	applog(LOG_ERR, "Chip %d failed PLL lock", chip_id);
+	return false;
+}
+
+static bool set_pll_config(struct A1_chain *a1,
+			   int ref_clock_khz, int sys_clock_khz)
+{
+	/*
+	 * TODO: this is only an initial approach with binary adjusted
+	 * dividers and thus not exploiting the whole divider range.
+	 *
+	 * If required, the algorithm can be adapted to find the PLL
+	 * parameters after:
+	 *
+	 * sys_clk = (ref_clk * pll_fbdiv) / (pll_prediv * pll_postdiv)
+	 *
+	 * with a higher pll_postdiv being desired over a higher pll_prediv
+	 */
+
+	int i, n;
+	static uint8_t writereg[128] = { 0x00, 0x00, 0x21, 0x84 /*0x80*/, };
+	uint8_t pre_div = 2;
+	uint8_t post_div = 4;
+	uint32_t fb_div = (sys_clock_khz * pre_div * post_div) / ref_clock_khz;
+
+	applog(LOG_WARNING, "Setting PLL: CLK_REF=%dMHz, SYS_CLK=%dMHz",
+	       ref_clock_khz / 1000, sys_clock_khz / 1000);
+
+	while (fb_div > 511) {
+		post_div <<= 1;
+		fb_div >>= 1;
+	}
+	if (post_div > 16) {
+		applog(LOG_WARNING, "Can't set PLL parameters");
+		return false;
+	}
+	writereg[0] = (pre_div << 6) | (post_div << 1) | (fb_div >> 8);
+	writereg[1] = fb_div & 0xff;
+	applog(LOG_WARNING, "Setting PLL to pre_div=%d, post_div=%d, fb_div=%d"
+	       ": 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x",
+	       pre_div, post_div, fb_div,
+	       writereg[0], writereg[1], writereg[2],
+	       writereg[3], writereg[4], writereg[5]);
+
+	if (!cmd_WRITE_REG_BCAST(a1, writereg))
+		return false;
+
+	for (i = 0; i < a1->num_active_chips; i++) {
+		int chip_id = i + 1;
+		if (!check_chip_pll_lock(a1, chip_id, writereg)) {
+			applog(LOG_ERR, "Chip %d failed PLL lock", chip_id);
+			return false;
+		}
+	}
+	return true;
+}
+
+/*
+ * BIST_START works only once after HW reset, on subsequent calls it
+ * returns 0 as number of chips.
+ *
+ * During testing / chip-bring up one might not be able or not want to reset
+ * chips on each cgminer restart, therefore this is an indirect alternative
+ * to enumerate chips in a chain by reading the chips' registers.
+ *
+ * TODO: to be removed after chip bring-up / testing
+ */
+static int manual_chain_detect(struct A1_chain *a1)
+{
+	spi_send_command(a1, A1_BIST_START, 0x00, NULL, 2);
+
+	int i;
+	int max_poll_words = MAX_CHAIN_LENGTH * 2;
+	for(i = 1; i < max_poll_words; i++) {
+		if (a1->spi_rx[0] == A1_BIST_START && a1->spi_rx[1] == 0) {
+			spi_transfer(a1->spi_ctx, NULL, a1->spi_rx, 2);
+			hexdump("RX", a1->spi_rx, 2);
+			uint8_t n = a1->spi_rx[1];
+			a1->num_chips = i / 2;
+			if (a1->num_chips != n) {
+				applog(LOG_ERR, "WARN: Enumeration: %d <-> %d",
+				       a1->num_chips, n);
+				if (n != 0)
+					a1->num_chips = n;
+			}
+			applog(LOG_WARNING, "manually detected %d chips", a1->num_chips);
+			return a1->num_chips;
+		}
+		bool s = spi_transfer(a1->spi_ctx, NULL, a1->spi_rx, 2);
+		hexdump("RX", a1->spi_rx, 2);
+		if (!s)
+			return false;
+	}
+	applog(LOG_WARNING, "Failed to manual find chips");
+	return false;
+}
+
+/********** disable / re-enable related section (temporary for testing) */
+static int get_current_ms(void)
+{
+	cgtimer_t ct;
+	cgtimer_time(&ct);
+	return cgtimer_to_ms(&ct);
+}
+
+static bool is_chip_disabled(struct A1_chain *a1, uint8_t chip_id)
+{
+	struct A1_chip *chip = &a1->chips[chip_id - 1];
+	return chip->cooldown_begin != 0;
+}
+
+/* check and disable chip, remember time */
+static void disable_chip(struct A1_chain *a1, uint8_t chip_id)
+{
+	struct A1_chip *chip = &a1->chips[chip_id - 1];
+	if (is_chip_disabled(a1, chip_id)) {
+		applog(LOG_WARNING, "Chip %d already disabled", chip_id);
+		return;
+	}
+	if (cmd_READ_REG(a1, chip_id)) {
+		applog(LOG_WARNING, "Chip %d is working, not going to disable",
+		       chip_id);
+		return;
+	}
+	applog(LOG_WARNING, "Disabling chip %d", chip_id);
+	chip->cooldown_begin = get_current_ms();
+}
+
+/* check if disabled chips can be re-enabled */
+void check_disabled_chips(struct A1_chain *a1)
+{
+	int i;
+	for (i = 0; i < a1->num_active_chips; i++) {
+		int chip_id = i + 1;
+		struct A1_chip *chip = &a1->chips[i];
+		if (!is_chip_disabled(a1, chip_id))
+			continue;
+		if (chip->cooldown_begin + COOLDOWN_MS > get_current_ms())
+			continue;
+		if (!cmd_READ_REG(a1, chip_id)) {
+			applog(LOG_WARNING, "Chip %d not yet working", chip_id);
+			/* restart cooldown period */
+			chip->cooldown_begin = get_current_ms();
+			continue;
+		}
+		applog(LOG_WARNING, "Chip %d is working again", chip_id);
+		chip->cooldown_begin = 0;
+	}
+}
+
+/********** job creation and result evaluation */
+static uint8_t *create_job(uint8_t chip_id, uint8_t job_id,
+			   const char *midstate, const char *wdata)
+{
+	static uint8_t job[WRITE_JOB_LENGTH] = {
+		/* command */
+		0x00, 0x00,
+		/* midstate */
+		0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+		0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+		0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+		0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+		/* wdata */
+		0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+		0x00, 0x00, 0x00, 0x00,
+		/* start nonce */
+		0x00, 0x00, 0x00, 0x00,
+		/* difficulty 1 */
+		0xff, 0xff, 0x00, 0x1d,
+		/* end nonce */
+		0xff, 0xff, 0xff, 0xff,
+	};
+	uint32_t *p1 = (uint32_t *) &job[34];
+	uint32_t *p2 = (uint32_t *) wdata;
+	unsigned char mid[32], data[12];
+
+	job[0] = (job_id << 4) | A1_WRITE_JOB;
+	job[1] = chip_id;
+
+	swab256(job + 2, midstate);
+	p1 = (uint32_t *) &job[34];
+	p2 = (uint32_t *) wdata;
+	p1[0] = bswap_32(p2[0]);
+	p1[1] = bswap_32(p2[1]);
+	p1[2] = bswap_32(p2[2]);
+	return job;
+}
+
+/* set work for given chip, returns true if a nonce range was finished */
+static bool set_work(struct A1_chain *a1, uint8_t chip_id, struct work *work)
+{
+	unsigned char *midstate = work->midstate;
+	unsigned char *wdata = work->data + 64;
+
+	struct A1_chip *chip = &a1->chips[chip_id - 1];
+	bool retval = false;
+
+	chip->last_queued_id++;
+	chip->last_queued_id &= 3;
+
+	if (chip->work[chip->last_queued_id] != NULL) {
+		work_completed(a1->cgpu, chip->work[chip->last_queued_id]);
+		chip->work[chip->last_queued_id] = NULL;
+		retval = true;
+	}
+	uint8_t *jobdata = create_job(chip_id, chip->last_queued_id + 1,
+				      midstate, wdata);
+	if (!cmd_WRITE_JOB(a1, chip_id, jobdata)) {
+		/* give back work */
+		work_completed(a1->cgpu, work);
+
+		applog(LOG_ERR, "Failed to set work for chip %d.%d",
+		       chip_id, chip->last_queued_id + 1);
+		// TODO: what else?
+	} else {
+		chip->work[chip->last_queued_id] = work;
+	}
+	return retval;
+}
+
+/* check for pending results in a chain, returns false if output queue empty */
+static bool get_nonce(struct A1_chain *a1, uint8_t *nonce,
+		      uint8_t *chip, uint8_t *job_id)
+{
+	int i;
+	if (!spi_send_command(a1, A1_READ_RESULT, 0x00, NULL, 0))
+		return false;
+
+	int max_poll_words = a1->num_chips * 2 + 1;
+	for(i = 0; i < max_poll_words; i++) {
+		if (!spi_transfer(a1->spi_ctx, NULL, a1->spi_rx, 2))
+			return false;
+		hexdump("RX", a1->spi_rx, 2);
+		if (a1->spi_rx[0] == A1_READ_RESULT && a1->spi_rx[1] == 0x00) {
+			applog(LOG_DEBUG, "Output queue empty");
+			return false;
+		}
+		if ((a1->spi_rx[0] & 0x0f) == A1_READ_RESULT &&
+		    a1->spi_rx[0] != 0) {
+			*job_id = a1->spi_rx[0] >> 4;
+			*chip = a1->spi_rx[1];
+
+			if (!spi_transfer(a1->spi_ctx, NULL, nonce, 4))
+				return false;
+
+			applog(LOG_DEBUG, "Got nonce for chip %d / job_id %d",
+			       *chip, *job_id);
+
+			return true;
+		}
+	}
+	applog(LOG_ERR, "Failed to poll for results");
+	return false;
+}
+
+/* reset input work queues in chip chain */
+static bool abort_work(struct A1_chain *a1)
+{
+	/*
+	 * TODO: implement reliable abort method
+	 * NOTE: until then, we are completing queued work => stales
+	 */
+	return true;
+}
+
+/********** driver interface */
+void exit_A1_chain(struct A1_chain *a1)
+{
+	if (a1 == NULL)
+		return;
+	free(a1->chips);
+	a1->chips = NULL;
+	spi_exit(a1->spi_ctx);
+	a1->spi_ctx = NULL;
+	free(a1);
+}
+
+
+struct A1_chain *init_A1_chain(struct spi_ctx *ctx)
+{
+	int i;
+	struct A1_chain *a1 = malloc(sizeof(*a1));
+	assert(a1 != NULL);
+
+	applog(LOG_DEBUG, "A1 init chain");
+	memset(a1, 0, sizeof(*a1));
+	a1->spi_ctx = ctx;
+
+	a1->num_chips = manual_chain_detect(a1);
+	if (a1->num_chips == 0)
+		goto failure;
+
+	applog(LOG_WARNING, "spidev%d.%d: Found %d A1 chips",
+	       a1->spi_ctx->config.bus, a1->spi_ctx->config.cs_line,
+	       a1->num_chips);
+
+	a1->num_active_chips = a1->num_chips;
+	if (config_options.override_chip_num > 0 &&
+	    a1->num_chips > config_options.override_chip_num) {
+		a1->num_active_chips = config_options.override_chip_num;
+		applog(LOG_WARNING, "Limiting chain to %d chips",
+				a1->num_active_chips);
+	}
+
+
+	if (!set_pll_config(a1, config_options.ref_clk_khz,
+			    config_options.sys_clk_khz))
+		goto failure;
+
+	a1->chips = calloc(a1->num_active_chips, sizeof(struct A1_chip));
+	assert (a1->chips != NULL);
+
+	if (!cmd_BIST_FIX_BCAST(a1))
+		goto failure;
+
+	for (i = 0; i < a1->num_active_chips; i++) {
+		int chip_id = i + 1;
+		if (!cmd_READ_REG(a1, chip_id)) {
+			applog(LOG_WARNING, "Failed to read register for "
+			       "chip %d -> disabling", chip_id);
+			a1->chips[i].num_cores = 0;
+			continue;
+		}
+		a1->chips[i].num_cores = a1->spi_rx[7];
+		a1->num_cores += a1->chips[i].num_cores;
+		applog(LOG_WARNING, "Found chip %d with %d active cores",
+		       chip_id, a1->chips[i].num_cores);
+	}
+	applog(LOG_WARNING, "Found %d chips with total %d active cores",
+	       a1->num_active_chips, a1->num_cores);
+
+	mutex_init(&a1->lock);
+	INIT_LIST_HEAD(&a1->active_wq.head);
+
+	return a1;
+
+failure:
+	exit_A1_chain(a1);
+	return NULL;
+}
+
+static bool A1_detect_one_chain(struct spi_config *cfg)
+{
+	struct cgpu_info *cgpu;
+	struct spi_ctx *ctx = spi_init(cfg);
+
+	if (ctx == NULL)
+		return false;
+
+	struct A1_chain *a1 = init_A1_chain(ctx);
+	if (a1 == NULL)
+		return false;
+
+	cgpu = malloc(sizeof(*cgpu));
+	assert(cgpu != NULL);
+
+	memset(cgpu, 0, sizeof(*cgpu));
+	cgpu->drv = &bitmineA1_drv;
+	cgpu->name = "BitmineA1";
+	cgpu->threads = 1;
+
+	cgpu->device_data = a1;
+
+	a1->cgpu = cgpu;
+	add_cgpu(cgpu);
+
+	return true;
+}
+
+#define MAX_SPI_BUS	1
+#define MAX_SPI_CS	2
+/* Probe SPI channel and register chip chain */
+void A1_detect(bool hotplug)
+{
+	int bus;
+	int cs_line;
+
+	/* no hotplug support for now */
+	if (hotplug)
+		return;
+
+	if (opt_bitmine_a1_options != NULL && parsed_config_options == NULL) {
+		int ref_clk;
+		int sys_clk;
+		int spi_clk;
+		int override_chip_num;
+
+		sscanf(opt_bitmine_a1_options, "%d:%d:%d:%d",
+		       &ref_clk, &sys_clk, &spi_clk,  &override_chip_num);
+		if (ref_clk != 0)
+			config_options.ref_clk_khz = ref_clk;
+		if (sys_clk != 0)
+			config_options.sys_clk_khz = sys_clk;
+		if (spi_clk != 0)
+			config_options.spi_clk_khz = spi_clk;
+		if (override_chip_num != 0)
+			config_options.override_chip_num = override_chip_num;
+
+		/* config options are global, scan them once */
+		parsed_config_options = &config_options;
+	}
+
+	applog(LOG_DEBUG, "A1 detect");
+	A1_hw_reset();
+	for (bus = 0; bus < MAX_SPI_BUS; bus++) {
+		for (cs_line = 0; cs_line < MAX_SPI_CS; cs_line++) {
+			struct spi_config cfg = default_spi_config;
+			cfg.mode = SPI_MODE_1;
+			cfg.speed = config_options.spi_clk_khz * 1000;
+			cfg.bus = bus;
+			cfg.cs_line = cs_line;
+			A1_detect_one_chain(&cfg);
+		}
+	}
+}
+
+/* return value is nonces processed since previous call */
+static int64_t A1_scanwork(struct thr_info *thr)
+{
+	int i;
+	struct cgpu_info *cgpu = thr->cgpu;
+	struct A1_chain *a1 = cgpu->device_data;
+	int32_t nonce_ranges_processed = 0;
+
+	applog(LOG_DEBUG, "A1 running scanwork");
+
+	uint32_t nonce;
+	uint8_t chip_id;
+	uint8_t job_id;
+	bool work_updated = false;
+
+	mutex_lock(&a1->lock);
+
+	/* poll queued results */
+	while (get_nonce(a1, (uint8_t*)&nonce, &chip_id, &job_id)) {
+		nonce = bswap_32(nonce);
+		work_updated = true;
+		if (chip_id < 1 || chip_id > a1->num_active_chips) {
+			applog(LOG_WARNING, "wrong chip_id %d",
+			       chip_id);
+			continue;
+		}
+		assert(job_id > 0 && job_id <= 4);
+		if (job_id < 1 || job_id > 4) {
+			applog(LOG_WARNING, "chip_id %d: wrong job_id %d",
+			       chip_id, job_id);
+			continue;
+		}
+
+		struct A1_chip *chip = &a1->chips[chip_id - 1];
+		struct work *work = chip->work[job_id - 1];
+		if (work == NULL) {
+			/* already been flushed => stale */
+			applog(LOG_WARNING, "chip %d: stale nonce 0x%08x",
+			       chip_id, nonce);
+			chip->stales++;
+			continue;
+		}
+		if (!submit_nonce(thr, work, nonce)) {
+			applog(LOG_WARNING, "chip %d: invalid nonce 0x%08x",
+			       chip_id, nonce);
+			chip->hw_errors++;
+			/* add a penalty of a full nonce range on HW errors */
+			nonce_ranges_processed--;
+			continue;
+		}
+		applog(LOG_WARNING, "YEAH: chip %d / job_id %d: nonce 0x%08x",
+		       chip_id, job_id, nonce);
+		chip->nonces_found++;
+	}
+
+	/* check for completed works */
+	for (i = 0; i < a1->num_active_chips; i++) {
+		uint8_t c = i + 1;
+		if (is_chip_disabled(a1, c))
+			continue;
+		if (!cmd_READ_REG(a1, c)) {
+			applog(LOG_ERR, "Failed to read reg from chip %d", c);
+			// TODO: what to do now?
+			disable_chip(a1, c);
+			continue;
+		}
+		if ((a1->spi_rx[5] & 0x02) != 0)
+			continue;
+
+		work_updated = true;
+		struct work *work;
+		work = wq_dequeue(&a1->active_wq);
+		if (work == NULL) {
+			applog(LOG_ERR, "Chip %d: work underflow", c);
+			break;
+		}
+		if (!set_work(a1, c, work))
+			continue;
+
+		nonce_ranges_processed++;
+		struct A1_chip *chip = &a1->chips[i];
+		chip->nonce_ranges_done++;
+		applog(LOG_DEBUG, "chip %d: job done => %d/%d/%d/%d", c,
+		       chip->nonce_ranges_done, chip->nonces_found,
+		       chip->hw_errors, chip->stales);
+
+		// repeat twice in case both jobs were done
+		if (!cmd_READ_REG(a1, i + 1)) {
+			applog(LOG_ERR, "Failed to read reg from chip %d", c);
+			// TODO: what to do now?
+			disable_chip(a1, c);
+			continue;
+		}
+		if ((a1->spi_rx[5] & 0x02) != 0)
+			continue;
+
+		work = wq_dequeue(&a1->active_wq);
+		if (work == NULL) {
+			applog(LOG_ERR, "Chip %d: work underflow", c);
+			break;
+		}
+		if (!set_work(a1, i + 1, work))
+			continue;
+
+		nonce_ranges_processed++;
+		chip->nonce_ranges_done++;
+		applog(LOG_DEBUG, "chip %d: job done => %d/%d/%d/%d",
+		       i + 1,
+		       chip->nonce_ranges_done, chip->nonces_found,
+		       chip->hw_errors, chip->stales);
+	}
+	check_disabled_chips(a1);
+	mutex_unlock(&a1->lock);
+
+	if (nonce_ranges_processed < 0) {
+		applog(LOG_ERR, "Negative nonce_processed");
+		nonce_ranges_processed = 0;
+	}
+
+	if (nonce_ranges_processed != 0) {
+		applog(LOG_DEBUG, "nonces processed %d",
+		       nonce_ranges_processed);
+	}
+	/* in case of no progress, prevent busy looping */
+	if (!work_updated)
+		cgsleep_ms(100);
+
+	return (int64_t)nonce_ranges_processed << 32;
+}
+
+/* queue two work items per chip in chain */
+static bool A1_queue_full(struct cgpu_info *cgpu)
+{
+	struct A1_chain *a1 = cgpu->device_data;
+	int queue_full = false;
+	struct work *work;
+
+	mutex_lock(&a1->lock);
+	applog(LOG_DEBUG, "A1 running queue_full: %d/%d",
+	       a1->active_wq.num_elems, a1->num_active_chips);
+
+	if (a1->active_wq.num_elems >= a1->num_active_chips * 2) {
+		applog(LOG_DEBUG, "active_wq full");
+		queue_full = true;
+	} else {
+		wq_enqueue(&a1->active_wq, get_queued(cgpu));
+	}
+	mutex_unlock(&a1->lock);
+
+	return queue_full;
+}
+
+static void A1_flush_work(struct cgpu_info *cgpu)
+{
+	struct A1_chain *a1 = cgpu->device_data;
+	applog(LOG_DEBUG, "A1 running flushwork");
+
+	int i;
+
+	mutex_lock(&a1->lock);
+	/* stop chips hashing current work */
+	if (!abort_work(a1)) {
+		applog(LOG_ERR, "failed to abort work in chip chain!");
+	}
+	/* flush the work chips were currently hashing */
+	for (i = 0; i < a1->num_active_chips; i++) {
+		int j;
+		struct A1_chip *chip = &a1->chips[i];
+		for (j = 0; j < 4; j++) {
+			struct work *work = chip->work[j];
+			if (work == NULL)
+				continue;
+			applog(LOG_DEBUG, "flushing chip %d, work %d: 0x%p",
+			       i, j + 1, work);
+			work_completed(cgpu, work);
+			chip->work[j] = NULL;
+		}
+	}
+	/* flush queued work */
+	applog(LOG_DEBUG, "flushing queued work...");
+	while (a1->active_wq.num_elems > 0) {
+		struct work *work = wq_dequeue(&a1->active_wq);
+		assert(work != NULL);
+		applog(LOG_DEBUG, "flushing 0x%p", work);
+		work_completed(cgpu, work);
+	}
+	mutex_unlock(&a1->lock);
+}
+
+static bool A1_thread_init(struct thr_info *t)
+{
+	return true;
+}
+static void A1_thread_shutdown(struct thr_info *t)
+{
+	if (log_file != NULL)
+		fclose(log_file);
+	log_file = NULL;
+}
+struct device_drv bitmineA1_drv = {
+	.drv_id = DRIVER_bitmineA1,
+	.dname = "BitmineA1",
+	.name = "BA1",
+	.drv_detect = A1_detect,
+	.thread_init = A1_thread_init,
+	.thread_shutdown = A1_thread_shutdown,
+
+	.hash_work = hash_queued_work,
+	.scanwork = A1_scanwork,
+	.queue_full = A1_queue_full,
+	.flush_work = A1_flush_work,
+};
diff --git a/miner.h b/miner.h
index 872365f..afd99e7 100644
--- a/miner.h
+++ b/miner.h
@@ -240,6 +240,7 @@ static inline int fsync (int fd)
 	DRIVER_ADD_COMMAND(hashfast) \
 	DRIVER_ADD_COMMAND(klondike) \
 	DRIVER_ADD_COMMAND(knc) \
+	DRIVER_ADD_COMMAND(bitmineA1) \
 	DRIVER_ADD_COMMAND(drillbit) \
 	DRIVER_ADD_COMMAND(bab) \
 	DRIVER_ADD_COMMAND(minion) \
@@ -982,6 +983,9 @@ extern char *opt_drillbit_options;
 #ifdef USE_BAB
 extern char *opt_bab_options;
 #endif
+#ifdef USE_BITMINE_A1
+extern char *opt_bitmine_a1_options;
+#endif
 #ifdef USE_USBUTILS
 extern char *opt_usb_select;
 extern int opt_usbdump;
@@ -1319,7 +1323,7 @@ struct work {
 	char		getwork_mode;
 };
 
-#ifdef USE_MODMINER 
+#ifdef USE_MODMINER
 struct modminer_fpga_state {
 	bool work_running;
 	struct work running_work;
diff --git a/spi-context.c b/spi-context.c
new file mode 100644
index 0000000..d070de2
--- /dev/null
+++ b/spi-context.c
@@ -0,0 +1,93 @@
+/*
+ * generic SPI functions
+ *
+ * Copyright 2013, 2014 Zefir Kurtisi <zefir.kurtisi@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 3 of the License, or (at your option)
+ * any later version.  See COPYING for more details.
+ */
+
+#include "spi-context.h"
+
+#include "logging.h"
+#include "miner.h"
+
+#include <fcntl.h>
+#include <sys/ioctl.h>
+#include <assert.h>
+
+struct spi_ctx *spi_init(struct spi_config *config)
+{
+	char dev_fname[PATH_MAX];
+	struct spi_ctx *ctx;
+
+	if (config == NULL)
+		return NULL;
+
+	sprintf(dev_fname, SPI_DEVICE_TEMPLATE, config->bus, config->cs_line);
+
+	int fd = open(dev_fname, O_RDWR);
+	if (fd < 0) {
+		applog(LOG_ERR, "SPI: Can not open SPI device %s", dev_fname);
+		return NULL;
+	}
+
+	if ((ioctl(fd, SPI_IOC_WR_MODE, &config->mode) < 0) ||
+	    (ioctl(fd, SPI_IOC_RD_MODE, &config->mode) < 0) ||
+	    (ioctl(fd, SPI_IOC_WR_BITS_PER_WORD, &config->bits) < 0) ||
+	    (ioctl(fd, SPI_IOC_RD_BITS_PER_WORD, &config->bits) < 0) ||
+	    (ioctl(fd, SPI_IOC_WR_MAX_SPEED_HZ, &config->speed) < 0) ||
+	    (ioctl(fd, SPI_IOC_RD_MAX_SPEED_HZ, &config->speed) < 0)) {
+		applog(LOG_ERR, "SPI: ioctl error on SPI device %s", dev_fname);
+		close(fd);
+		return NULL;
+	}
+
+	ctx = malloc(sizeof(*ctx));
+	assert(ctx != NULL);
+
+	ctx->fd = fd;
+	ctx->config = *config;
+	applog(LOG_WARNING, "SPI '%s': mode=%hhu, bits=%hhu, speed=%u",
+	       dev_fname, ctx->config.mode, ctx->config.bits,
+	       ctx->config.speed);
+	return ctx;
+}
+
+extern void spi_exit(struct spi_ctx *ctx)
+{
+	if (NULL == ctx)
+		return;
+
+	close(ctx->fd);
+	free(ctx);
+}
+
+extern bool spi_transfer(struct spi_ctx *ctx, uint8_t *txbuf,
+			 uint8_t *rxbuf, int len)
+{
+	struct spi_ioc_transfer xfr;
+	int ret;
+
+	if (rxbuf != NULL)
+		memset(rxbuf, 0xff, len);
+
+	ret = len;
+
+	xfr.tx_buf = (unsigned long)txbuf;
+	xfr.rx_buf = (unsigned long)rxbuf;
+	xfr.len = len;
+	xfr.speed_hz = ctx->config.speed;
+	xfr.delay_usecs = ctx->config.delay;
+	xfr.bits_per_word = ctx->config.bits;
+	xfr.cs_change = 0;
+	xfr.pad = 0;
+
+	ret = ioctl(ctx->fd, SPI_IOC_MESSAGE(1), &xfr);
+	if (ret < 1)
+		applog(LOG_ERR, "SPI: ioctl error on SPI device: %d", ret);
+
+	return ret > 0;
+}
diff --git a/spi-context.h b/spi-context.h
new file mode 100644
index 0000000..a341e69
--- /dev/null
+++ b/spi-context.h
@@ -0,0 +1,48 @@
+#ifndef SPI_CONTEXT_H
+#define SPI_CONTEXT_H
+
+#include <linux/types.h>
+#include <linux/spi/spidev.h>
+#include <stdbool.h>
+#include <stdint.h>
+
+#define SPI_DEVICE_TEMPLATE		"/dev/spidev%d.%d"
+#define DEFAULT_SPI_BUS			0
+#define DEFAULT_SPI_CS_LINE		0
+#define DEFAULT_SPI_MODE		SPI_MODE_0
+#define DEFAULT_SPI_BITS_PER_WORD	8
+#define DEFAULT_SPI_SPEED		1500000
+#define DEFAULT_SPI_DELAY_USECS		0
+
+struct spi_config {
+	int bus;
+	int cs_line;
+	uint8_t mode;
+	uint32_t speed;
+	uint8_t bits;
+	uint16_t delay;
+};
+
+static const struct spi_config default_spi_config = {
+	.bus		= DEFAULT_SPI_BUS,
+	.cs_line	= DEFAULT_SPI_CS_LINE,
+	.mode		= DEFAULT_SPI_MODE,
+	.speed		= DEFAULT_SPI_SPEED,
+	.bits		= DEFAULT_SPI_BITS_PER_WORD,
+	.delay		= DEFAULT_SPI_DELAY_USECS,
+};
+
+struct spi_ctx {
+	int fd;
+	struct spi_config config;
+};
+
+/* create SPI context with given configuration, returns NULL on failure */
+extern struct spi_ctx *spi_init(struct spi_config *config);
+/* close descriptor and free resources */
+extern void spi_exit(struct spi_ctx *ctx);
+/* process RX/TX transfer, ensure buffers are long enough */
+extern bool spi_transfer(struct spi_ctx *ctx, uint8_t *txbuf,
+			 uint8_t *rxbuf, int len);
+
+#endif /* SPI_CONTEXT_H */