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
|
/*
* hsm.c
* ----------------
* Main module for the HSM project.
*
* Copyright (c) 2016-2017, NORDUnet A/S All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
* - Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
*
* - Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* - Neither the name of the NORDUnet nor the names of its contributors may
* be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
* IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
* TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
* PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
* TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
* LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/*
* This is the main RPC server module. At the moment, it has a single
* worker thread to handle RPC requests, while the main thread handles CLI
* activity. The design allows for multiple worker threads to handle
* concurrent RPC requests from multiple clients (muxed through a daemon
* on the host).
*/
#include <string.h>
/* Rename both CMSIS HAL_OK and libhal HAL_OK to disambiguate */
#define HAL_OK CMSIS_HAL_OK
#include "cmsis_os.h"
#include "stm-init.h"
#include "stm-led.h"
#include "stm-fmc.h"
#include "stm-uart.h"
#include "stm-sdram.h"
#include "mgmt-cli.h"
#undef HAL_OK
#define HAL_OK LIBHAL_OK
#include "hal.h"
#include "hal_internal.h"
#include "slip_internal.h"
#include "xdr_internal.h"
#undef HAL_OK
#ifndef NUM_RPC_TASK
/* Just one RPC task for now. More will require active resource management
* of at least the FPGA cores.
*/
#define NUM_RPC_TASK 1
#endif
#ifndef TASK_STACK_SIZE
/* Define an absurdly large task stack, because some pkey operation use a
* lot of stack variables. This has to go in SDRAM, because it exceeds the
* total RAM on the ARM.
*/
#define TASK_STACK_SIZE 200*1024
#endif
#ifndef MAX_PKT_SIZE
/* An arbitrary number, more or less driven by the 4096-bit RSA
* keygen test.
*/
#define MAX_PKT_SIZE 4096
#endif
/* RPC buffers. For each active RPC, there will be two - input and output.
*/
typedef struct {
size_t len;
uint8_t buf[MAX_PKT_SIZE];
} rpc_buffer_t;
/* A mail queue (memory pool + message queue) for RPC request messages.
*/
osMailQId ibuf_queue;
osMailQDef(ibuf_queue, NUM_RPC_TASK + 2, rpc_buffer_t);
#if NUM_RPC_TASK > 1
/* A mutex to arbitrate concurrent UART transmits, from RPC responses.
*/
osMutexId uart_mutex;
osMutexDef(uart_mutex);
static inline void uart_lock(void) { osMutexWait(uart_mutex, osWaitForever); }
static inline void uart_unlock(void) { osMutexRelease(uart_mutex); }
#else
static inline void uart_lock(void) { }
static inline void uart_unlock(void) { }
#endif
#if NUM_RPC_TASK > 1
/* A mutex to arbitrate concurrent access to the keystore.
*/
osMutexId ks_mutex;
osMutexDef(ks_mutex);
void hal_ks_lock(void) { osMutexWait(ks_mutex, osWaitForever); }
void hal_ks_unlock(void) { osMutexRelease(ks_mutex); }
#endif
static uint8_t uart_rx[2]; /* current character received from UART */
static uint32_t uart_rx_idx = 0;
/* Callback for HAL_UART_Receive_DMA().
* With multiple worker threads, we can't do a blocking receive, because
* that prevents other threads from sending RPC responses (because they
* both want to lock the UART - see stm32f4xx_hal_uart.c). So we have to
* do a non-blocking receive with a callback routine.
* Even with only one worker thread, context-switching to the CLI thread
* causes HAL_UART_Receive to miss input.
*/
static void RxCallback(uint8_t c)
{
/* current RPC input buffer */
static rpc_buffer_t *ibuf = NULL;
int complete;
if (ibuf == NULL) {
if ((ibuf = (rpc_buffer_t *)osMailAlloc(ibuf_queue, 0)) == NULL)
/* This could happen if all dispatch threads are busy, and
* there are NUM_RPC_TASK requests already queued. We'd like
* to to send a "server busy" error, but we've just received
* the first byte of the request, so we don't yet have enough
* context to craft a response.
*/
return;
ibuf->len = 0;
}
if (hal_slip_process_char(c, ibuf->buf, &ibuf->len, sizeof(ibuf->buf), &complete) != LIBHAL_OK)
Error_Handler();
if (complete) {
if (osMailPut(ibuf_queue, (void *)ibuf) != osOK)
Error_Handler();
ibuf = NULL;
}
}
void HAL_UART2_RxHalfCpltCallback(UART_HandleTypeDef *huart)
{
RxCallback(uart_rx[uart_rx_idx]);
uart_rx_idx ^= 1;
}
void HAL_UART2_RxCpltCallback(UART_HandleTypeDef *huart)
{
RxCallback(uart_rx[uart_rx_idx]);
uart_rx_idx ^= 1;
}
void HAL_UART_ErrorCallback(UART_HandleTypeDef *huart)
{
/* I dunno, just trap it for now */
Error_Handler();
}
hal_error_t hal_serial_send_char(uint8_t c)
{
return (uart_send_char2(STM_UART_USER, c) == 0) ? LIBHAL_OK : HAL_ERROR_RPC_TRANSPORT;
}
/* Thread entry point for the RPC request handler.
*/
void dispatch_thread(void const *args)
{
rpc_buffer_t obuf_s, *obuf = &obuf_s, *ibuf;
hal_crc32_t crc;
while (1) {
memset(obuf, 0, sizeof(*obuf));
obuf->len = sizeof(obuf->buf);
/* Wait for a complete RPC request */
osEvent evt = osMailGet(ibuf_queue, osWaitForever);
if (evt.status != osEventMail)
continue;
ibuf = (rpc_buffer_t *)evt.value.p;
if (ibuf->len < 8)
continue;
crc = hal_crc32_init();
/* Calculate CRC32 checksum of the contents, after SLIP decoding */
crc = hal_crc32_update(crc, ibuf->buf, ibuf->len);
crc = hal_crc32_finalize(crc);
if (crc != 0xffffffff) {
/* XXX Full-stop on CRC errors is probably not the best long-term solution,
* but it helps while we are sorting out any remaining issues.
*/
led_on(LED_RED);
/* Steal UART lock to stop all threads */
uart_lock();
Error_Handler();
}
/* Remove CRC from end of ibuf->buf */
ibuf->len -= 4;
ibuf->buf[ibuf->len] = 0xc0;
/* Process the request */
hal_error_t ret = hal_rpc_server_dispatch(ibuf->buf, ibuf->len, obuf->buf, &obuf->len);
osMailFree(ibuf_queue, (void *)ibuf);
if (ret != LIBHAL_OK) {
/* If hal_rpc_server_dispatch failed with an XDR error, it
* probably means the request packet was garbage. In any case, we
* have nothing to transmit.
*/
continue;
}
/* Send the response */
uart_lock();
ret = hal_rpc_sendto(obuf->buf, obuf->len, NULL);
uart_unlock();
if (ret != LIBHAL_OK)
Error_Handler();
}
}
osThreadDef_t thread_def[NUM_RPC_TASK];
/* Allocate memory from SDRAM1. There is only malloc, no free, so we don't
* worry about fragmentation. */
static uint8_t *sdram_malloc(size_t size)
{
/* end of variables declared with __attribute__((section(".sdram1"))) */
extern uint8_t _esdram1 __asm ("_esdram1");
/* end of SDRAM1 section */
extern uint8_t __end_sdram1 __asm ("__end_sdram1");
static uint8_t *sdram_heap = &_esdram1;
uint8_t *p = sdram_heap;
#define pad(n) (((n) + 3) & ~3)
size = pad(size);
if (p + size > &__end_sdram1)
return NULL;
sdram_heap += size;
return p;
}
/* Implement static memory allocation for libhal over sdram_malloc().
* Once again, there's only alloc, not free. */
void *hal_allocate_static_memory(const size_t size)
{
return sdram_malloc(size);
}
#if NUM_RPC_TASK > 1
/* Critical section start/end, currently used just for hal_core_alloc/_free.
*/
void hal_critical_section_start(void)
{
__disable_irq();
}
void hal_critical_section_end(void)
{
__enable_irq();
}
#endif
/* The main thread. This does all the setup, and the worker threads handle
* the rest.
*/
int main()
{
stm_init();
uart_set_default(STM_UART_MGMT);
led_on(LED_GREEN);
/* Prepare FMC interface. */
fmc_init();
sdram_init();
if ((ibuf_queue = osMailCreate(osMailQ(ibuf_queue), NULL)) == NULL)
Error_Handler();
#if NUM_RPC_TASK > 1
if ((uart_mutex = osMutexCreate(osMutex(uart_mutex))) == NULL)
Error_Handler();
if ((ks_mutex = osMutexCreate(osMutex(ks_mutex))) == NULL)
Error_Handler();
#endif
if (hal_rpc_server_init() != LIBHAL_OK)
Error_Handler();
/* Create the rpc dispatch worker threads. */
for (int i = 0; i < NUM_RPC_TASK; ++i) {
osThreadDef_t *ot = &thread_def[i];
ot->pthread = dispatch_thread;
ot->tpriority = osPriorityNormal;
ot->stacksize = TASK_STACK_SIZE;
ot->stack_pointer = (uint32_t *)(sdram_malloc(TASK_STACK_SIZE));
if (ot->stack_pointer == NULL)
Error_Handler();
if (osThreadCreate(ot, (void *)i) == NULL)
Error_Handler();
}
/* Start the UART receiver. */
if (HAL_UART_Receive_DMA(&huart_user, uart_rx, 2) != CMSIS_HAL_OK)
Error_Handler();
/* Launch other threads (csprng warm-up thread?)
* Wait for FPGA_DONE interrupt.
*/
return cli_main();
}
|