August 2017 - devel - libdivecomputer.org

serial_ftdi Issues
by John Van Ostrand 29 Aug '17

29 Aug '17

I'm trying to get the Cochran computers working with the FTDI library and serial_ftdi so I can read them on an Android biuld. I'm having two issues that I have to ask for advice about. 1. In serial_ftdi_read() a failed read exponentially backs off when there's a short read. This is causing Cochran reads to timeout about 95% of the time. It's as if reading too late causes lost data. Having a fixed backoff of 10ms works. Are there any issues if I alter this backoff tactic? It could be shortened to 1ms if 10ms is too long. diff --git a/core/serial_ftdi.c b/core/serial_ftdi.c index f8cba95..bc20ce1 100644 --- a/core/serial_ftdi.c +++ b/core/serial_ftdi.c @@ -384,7 +384,7 @@ static dc_status_t serial_ftdi_read (dc_custom_io_t *io, void *data, size_t size if (timeout == -1) timeout = 10000; - int backoff = 1; + int backoff = 10; int slept = 0; unsigned int nbytes = 0; while (nbytes < size) { @@ -402,12 +402,11 @@ static dc_status_t serial_ftdi_read (dc_custom_io_t *io, void *data, size_t size } serial_ftdi_sleep (device, backoff); slept += backoff; - backoff *= 2; if (backoff + slept > timeout) backoff = timeout - slept; } else { // Reset backoff to 1 on success. - backoff = 1; + backoff = 10; } nbytes += n; 2. It doesn't make any sense to me but setting serial break on then off is the trigger to the Cochran to wake up, so I need the ability to set breaks. Currently the code doesn't support setting breaks and is explained with the comment "// Not implemented in libftdi yet. Research it further." So I looked at it and it's supported with the following libftdi-1.3 function: int ftdi_set_line_property2(struct ftdi_context *ftdi, enum ftdi_bits_type bits, enum ftdi_stopbits_type sbit, enum ftdi_parity_type parity, enum ftdi_break_type break_type); To make this work seamlessly with libdivecomputer's serial function I need to emulate the dc_serial_set_break function. A prototype serial_ftdi_set_break function is there and I can change it to use ftdi_set_line_property2, which means I have to get the parameters bits, sbit, and parity from somewhere. There isn't a function that will read this from the device. I could create some global vars in serial_ftdi.c, initialize them with 8,1,N values, update them as needed. To support multiple devices concurrently it would have to be an array indexed by device but it's not clear what index would work best. I was thinking of a global variable like: struct serial_parm { struct ftdi_context *ftdi_ctx; enum ftdi_bits_type bits; enum ftdi_stopbits_type sbit; enum ftdi_parity_type parity; } *serial_parms = NULL; Then realloc() to add (on serial_ftdi_open) or remove (on serial_ftdi_close). When serial_ftdi_configure is called I can search the array and update the values so I can use them in serial_ftdi_set_break. Does this work or is there a better idea? -- John Van Ostrand At large on sabbatical

1 0

Backwards incompatible changes
by Jef Driesen 18 Aug '17

18 Aug '17

Hi, I made two small backwards incompatible changes to the public api: The dc_datetime_t structure has been extended with a new "timezone" field, containing the UTC offset in seconds. Devices without timezone support will set the field to the special value DC_TIMEZONE_NONE. The hw_{frog,ostc,ostc3}_device_clock() functions have been removed in favor of the new dc_device_timesync() function. Jef

1 0

[RFD] divecomputer time/date synchronization while downloading dives
by Linus Torvalds 18 Aug '17

18 Aug '17

I'm slightly OCD when it comes to clocks. My wrist watch is one of those "radio time signal" ones. Seeing wall clocks that are the wrong time drives me wild. I just can't take it. "You have *one* job, and you suck at it". When I was at the conference hotel in Beijing last month, I reacted to how almost *every* clock I saw was off by minutes, sometimes hours. The time display in the elevator? Off by a lot. The time in the room thermostat? Completely random. So guess what happens with dive computers? I react to them when they have clocks that run slow or fast. My old Uemis would be off by minutes when I hadn't used it in a couple of months, and it was annoying to set the right time, and boy did I notice. Anyway, what this all builds up to is that I'd actually like to just set the time automatically when I connect the dive computer to my laptop (or my cellphone, for that matter - usually those end up being synchronized to the local time quite nicely). And I just wrote the code for that for the Suunto EON Steel. Patch to libdivecomputer attached in case anybody cares (and I was happy to realize that it actually uses milliseconds for synchronization, but then noticed that it may be just a protocol thing, and the extra precision may get thrown away). But I suspect people will hate that patch because it does that time sync thing *unconditionally*. It would be good to have a check-box and some way to actually set the "sync time" flag. But for that to happen, we'd need not just the UI side, we'd need to have the libdivecomputer interface too. And yes, I know about "configure dive computer" and the "sync time" hack there. But it's really an ostc-specific hack, not something that can be used in general to synchronize the clock. Thus the "request for discussion". I'd love to be able to pass in flags like this to dc_device_open() or add a sane "dc_device_set_feature()" interface or something.. Linus

5 8

Reading log files
by John Van Ostrand 18 Aug '17

18 Aug '17

Should libdivecomputer be able to read other sources of dives other than dive computers? It seems to me that the interface is suitable for reading a dive log file. The idea would be that instead of passing LDC a serial file, you give it a log file name and instead of a family, model and device you give it a manufacturer, software and file type. LDC opens the file, decodes what it needs to, and starts passing dives to the callback function. I don't know about other software logs but the Cochran Analyst logs that I've been working with basically stores the log and dive data in the exact same format as was retrieved from the DC. So the parse functions are probably identical. -- John Van Ostrand At large on sabbatical

2 1

[PATCH 1/2] Cochran: Profile buffer size fix and address size fix
by John Van Ostrand 14 Aug '17

14 Aug '17

I dived the model enough to wrap the profile buffer and I was wrong about where the end was. Also, the buffer starts 3 bytes after where it could. We were treating profile pointers as 4 bytes when they are two bytes. This worked most of the time when short tissues were clear (tissue load follows the pointer). --- src/cochran_commander.c | 28 +++++++++++++++++++--------- 1 file changed, 19 insertions(+), 9 deletions(-) diff --git a/src/cochran_commander.c b/src/cochran_commander.c index 9670aea..921c8c0 100644 --- a/src/cochran_commander.c +++ b/src/cochran_commander.c @@ -127,11 +127,11 @@ static const cochran_device_layout_t cochran_cmdr_tm_device_layout = { 0xffffff, // cf_last_interdive 0x15c, // cf_serial_number 0x010000, // rb_logbook_begin - 0x012328, // rb_logbook_end + 0x01232b, // rb_logbook_end 90, // rb_logbook_entry_size 100, // rb_logbook_entry_count - 0x012328, // rb_profile_begin - 0x020000, // rb_profile_end + 0x01232b, // rb_profile_begin + 0x018000, // rb_profile_end 15, // pt_fingerprint 4, // fingerprint_size 0, // pt_profile_pre @@ -621,7 +621,7 @@ cochran_commander_find_fingerprint(cochran_commander_device_t *device, cochran_d unsigned int rb_head_ptr; if (device->layout->model == COCHRAN_MODEL_COMMANDER_TM) // TM uses SRAM and does not need to erase pages - rb_head_ptr = base + array_uint32_word_be(data->config + device->layout->cf_last_log); + rb_head_ptr = base + array_uint16_be(data->config + device->layout->cf_last_log); else if (device->layout->endian == ENDIAN_WORD_BE) rb_head_ptr = base + (array_uint32_word_be(data->config + device->layout->cf_last_log) & 0xfffff000) + 0x2000; else @@ -639,9 +639,16 @@ cochran_commander_find_fingerprint(cochran_commander_device_t *device, cochran_d } unsigned int last_profile_idx = (device->layout->rb_logbook_entry_count + head_dive - 1) % device->layout->rb_logbook_entry_count; - unsigned int last_profile_end = base + array_uint32_le(data->logbook + last_profile_idx * device->layout->rb_logbook_entry_size + device->layout->pt_profile_end); - unsigned int last_profile_pre = 0xFFFFFFFF; + unsigned int last_profile_end = 0; + if (device->layout->model == COCHRAN_MODEL_COMMANDER_TM) + // There is no end pointer in this model and no inter-dive + // events. We could use profile_begin from the next dive but + // since this is the last dive, we'll use rb_head_ptr + last_profile_end = rb_head_ptr; + else + last_profile_end = base + array_uint32_le(data->logbook + last_profile_idx * device->layout->rb_logbook_entry_size + device->layout->pt_profile_end); + unsigned int last_profile_pre = 0xFFFFFFFF; if (device->layout->endian == ENDIAN_WORD_BE) last_profile_pre = base + array_uint32_word_be(data->config + device->layout->cf_last_log); else @@ -663,7 +670,11 @@ cochran_commander_find_fingerprint(cochran_commander_device_t *device, cochran_d break; } - unsigned int profile_pre = base + array_uint32_le(log_entry + device->layout->pt_profile_pre); + unsigned int profile_pre = 0; + if (device->layout->model == COCHRAN_MODEL_COMMANDER_TM) + profile_pre = base + array_uint16_le(log_entry + device->layout->pt_profile_pre); + else + profile_pre = base + array_uint32_le(log_entry + device->layout->pt_profile_pre); unsigned int sample_size = cochran_commander_profile_size(device, data, i, profile_pre, last_profile_pre); last_profile_pre = profile_pre; @@ -931,7 +942,6 @@ cochran_commander_device_foreach (dc_device_t *abstract, dc_dive_callback_t call // Locate fingerprint, recent dive with invalid profile and calc read size unsigned int profile_read_size = cochran_commander_find_fingerprint(device, &data); - // Update progress indicator with new maximum progress.maximum -= (max_sample - profile_read_size); device_event_emit (abstract, DC_EVENT_PROGRESS, &progress); @@ -991,7 +1001,7 @@ cochran_commander_device_foreach (dc_device_t *abstract, dc_dive_callback_t call unsigned int sample_end_address = 0; if (layout->model == COCHRAN_MODEL_COMMANDER_TM) { - sample_start_address = base + array_uint24_le (log_entry + layout->pt_profile_begin); + sample_start_address = base + array_uint16_le (log_entry + layout->pt_profile_begin); sample_end_address = last_start_address; // Commander TM has SRAM which seems to randomize when they lose power for too long // Check for bad entries. -- 2.4.11

1 1

[PATCH] Cochran: Added three event codes
by John Van Ostrand 13 Aug '17

13 Aug '17

I found three previously undiscovered event codes. --- src/cochran_commander_parser.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/cochran_commander_parser.c b/src/cochran_commander_parser.c index 60a1b31..c3dad26 100644 --- a/src/cochran_commander_parser.c +++ b/src/cochran_commander_parser.c @@ -222,7 +222,9 @@ static const cochran_events_t cochran_events[] = { {0xA9, 1, SAMPLE_EVENT_SURFACE, SAMPLE_FLAGS_END}, // Exited PDI mode {0xAB, 5, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Ceiling decrease {0xAD, 5, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Ceiling increase + {0xB5, 1, SAMPLE_EVENT_AIRTIME, SAMPLE_FLAGS_BEGIN}, // Air < 5 mins deco {0xBD, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Switched to nomal PO2 setting + {0xBE, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Ceiling > 60 ft {0xC0, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Switched to FO2 21% mode {0xC1, 1, SAMPLE_EVENT_ASCENT, SAMPLE_FLAGS_BEGIN}, // Ascent rate greater than limit {0xC2, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Low battery warning @@ -233,6 +235,7 @@ static const cochran_events_t cochran_events[] = { {0xC8, 1, SAMPLE_EVENT_PO2, SAMPLE_FLAGS_BEGIN}, // PO2 too high {0xCC, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_BEGIN}, // Low Cylinder 1 pressure {0xCE, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_BEGIN}, // Non-decompression warning + {0xCF, 1, SAMPLE_EVENT_OLF, SAMPLE_FLAGS_BEGIN}, // O2 Toxicity {0xCD, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Switched to deco blend {0xD0, 1, SAMPLE_EVENT_WORKLOAD, SAMPLE_FLAGS_BEGIN}, // Breathing rate alarm {0xD3, 1, SAMPLE_EVENT_NONE, SAMPLE_FLAGS_NONE}, // Low gas 1 flow rate -- 2.4.11

1 0

[PATCH 1/6] Cochran: Remove uneeded cochran_commander_get_sample_parms function
by John Van Ostrand 11 Aug '17

11 Aug '17

Since moving to per-dive download of profile data (and now rbstream download) the data->sample_data_offset and data->sample_size variables aren't used so calculating them doesn't make sense. --- src/cochran_commander.c | 68 ------------------------------------------------- 1 file changed, 68 deletions(-) diff --git a/src/cochran_commander.c b/src/cochran_commander.c index e517027..3e125bb 100644 --- a/src/cochran_commander.c +++ b/src/cochran_commander.c @@ -59,9 +59,6 @@ typedef struct cochran_data_t { int invalid_profile_dive_num; unsigned int logbook_size; - - unsigned int sample_data_offset; - unsigned int sample_size; } cochran_data_t; typedef struct cochran_device_layout_t { @@ -608,69 +605,6 @@ cochran_commander_find_fingerprint(cochran_commander_device_t *device, cochran_d } - -static void -cochran_commander_get_sample_parms(cochran_commander_device_t *device, cochran_data_t *data) -{ - dc_device_t *abstract = (dc_device_t *) device; - unsigned int pre_dive_offset = 0, end_dive_offset = 0; - - unsigned int dive_count = 0; - if (data->dive_count < device->layout->rb_logbook_entry_count) - dive_count = data->dive_count; - else - dive_count = device->layout->rb_logbook_entry_count; - - // Find lowest and highest offsets into sample data - unsigned int low_offset = 0xFFFFFFFF; - unsigned int high_offset = 0; - - for (unsigned int i = data->fp_dive_num + 1; i < dive_count; i++) { - pre_dive_offset = array_uint32_le (data->logbook + i * device->layout->rb_logbook_entry_size - + device->layout->pt_profile_pre); - end_dive_offset = array_uint32_le (data->logbook + i * device->layout->rb_logbook_entry_size - + device->layout->pt_profile_end); - - // Validate offsets, allow 0xFFFFFFF for end_dive_offset - // because we handle that as a special case. - if (pre_dive_offset < device->layout->rb_profile_begin || - pre_dive_offset > device->layout->rb_profile_end) { - ERROR(abstract->context, "Invalid pre-dive offset (%08x) on dive %d.", pre_dive_offset, i); - continue; - } - - if (end_dive_offset < device->layout->rb_profile_begin || - (end_dive_offset > device->layout->rb_profile_end && - end_dive_offset != 0xFFFFFFFF)) { - ERROR(abstract->context, "Invalid end-dive offset (%08x) on dive %d.", end_dive_offset, i); - continue; - } - - // Check for ring buffer wrap-around. - if (pre_dive_offset > end_dive_offset) - break; - - if (pre_dive_offset < low_offset) - low_offset = pre_dive_offset; - if (end_dive_offset > high_offset && end_dive_offset != 0xFFFFFFFF ) - high_offset = end_dive_offset; - } - - if (pre_dive_offset > end_dive_offset) { - high_offset = device->layout->rb_profile_end; - low_offset = device->layout->rb_profile_begin; - data->sample_data_offset = low_offset; - data->sample_size = high_offset - low_offset; - } else if (low_offset < 0xFFFFFFFF && high_offset > 0) { - data->sample_data_offset = low_offset; - data->sample_size = high_offset - data->sample_data_offset; - } else { - data->sample_data_offset = 0; - data->sample_size = 0; - } -} - - dc_status_t cochran_commander_device_open (dc_device_t **out, dc_context_t *context, const char *name) { @@ -907,8 +841,6 @@ cochran_commander_device_foreach (dc_device_t *abstract, dc_dive_callback_t call progress.maximum -= (max_sample - profile_read_size); device_event_emit (abstract, DC_EVENT_PROGRESS, &progress); - cochran_commander_get_sample_parms(device, &data); - // Emit a device info event. dc_event_devinfo_t devinfo; devinfo.model = layout->model; -- 2.4.11

2 15

[PATCH 1/3] Cochran: Add a size parameter to cochran_commander_read_config().
by John Van Ostrand 11 Aug '17

11 Aug '17

Responded to a code review. --- src/cochran_commander.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/src/cochran_commander.c b/src/cochran_commander.c index f7d2e91..068fd04 100644 --- a/src/cochran_commander.c +++ b/src/cochran_commander.c @@ -429,12 +429,24 @@ cochran_commander_read_id (cochran_commander_device_t *device, unsigned char id[ static dc_status_t -cochran_commander_read_config (cochran_commander_device_t *device, dc_event_progress_t *progress, unsigned char data[]) +cochran_commander_read_config (cochran_commander_device_t *device, dc_event_progress_t *progress, unsigned char data[], int data_size) { dc_device_t *abstract = (dc_device_t *) device; dc_status_t rc = DC_STATUS_SUCCESS; - unsigned int pages = 2; + unsigned int pages = 0; + + switch (data_size) { + case 512: + pages = 1; + break; + case 1024: + pages = 2; + break; + default: + return DC_STATUS_INVALIDARGS; + break; + } if (device->layout->model == COCHRAN_MODEL_COMMANDER_TM) pages = 1; @@ -816,7 +828,7 @@ cochran_commander_device_dump (dc_device_t *abstract, dc_buffer_t *buffer) vendor.size = sizeof (device->id); device_event_emit (abstract, DC_EVENT_VENDOR, &vendor); - rc = cochran_commander_read_config (device, &progress, config); + rc = cochran_commander_read_config (device, &progress, config, config_size); if (rc != DC_STATUS_SUCCESS) return rc; @@ -867,7 +879,7 @@ cochran_commander_device_foreach (dc_device_t *abstract, dc_dive_callback_t call // Read config dc_status_t rc = DC_STATUS_SUCCESS; - rc = cochran_commander_read_config(device, &progress, data.config); + rc = cochran_commander_read_config(device, &progress, data.config, max_config); if (rc != DC_STATUS_SUCCESS) return rc; -- 2.4.11

1 2