]> sigrok.org Git - libsigrokdecode.git/log
libsigrokdecode.git
4 years agocan: decode CAN-FD DLC
Stephan Thiele [Fri, 22 Feb 2019 18:50:03 +0000 (19:50 +0100)]
can: decode CAN-FD DLC

4 years agocan: display CRC type + correct decoding of CRC-17
Stephan Thiele [Fri, 22 Feb 2019 18:29:53 +0000 (19:29 +0100)]
can: display CRC type + correct decoding of CRC-17

4 years agocan: implement decoding of CAN-FD header when FDF bit is set
Stephan Thiele [Wed, 20 Feb 2019 22:36:18 +0000 (23:36 +0100)]
can: implement decoding of CAN-FD header when FDF bit is set

4 years agocan: implement basic CAN-FD decoding (SRR and FDF bits)
Stephan Thiele [Thu, 10 Jan 2019 22:59:27 +0000 (23:59 +0100)]
can: implement basic CAN-FD decoding (SRR and FDF bits)

4 years agoeeprom93xx: Add ascii format support
Sylvain Pelissier [Fri, 19 Jul 2019 12:32:01 +0000 (14:32 +0200)]
eeprom93xx: Add ascii format support

4 years agomrf24j40: Add TX retries/fails and CCAFAIL support.
Uwe Hermann [Tue, 6 Aug 2019 19:42:07 +0000 (21:42 +0200)]
mrf24j40: Add TX retries/fails and CCAFAIL support.

4 years agomrf24j40: Add RX/TX frame annotations.
Uwe Hermann [Wed, 31 Jul 2019 19:40:07 +0000 (21:40 +0200)]
mrf24j40: Add RX/TX frame annotations.

4 years agopca9571: Shorten names a bit.
Uwe Hermann [Sat, 6 Jul 2019 13:33:43 +0000 (15:33 +0200)]
pca9571: Shorten names a bit.

4 years agoAdd an NXP PCA9571 decoder.
Mickael Bosch [Tue, 2 Jul 2019 14:31:12 +0000 (16:31 +0200)]
Add an NXP PCA9571 decoder.

4 years agomodbus: Make the 'framegap' option an integer.
Uwe Hermann [Sun, 30 Jun 2019 22:27:17 +0000 (00:27 +0200)]
modbus: Make the 'framegap' option an integer.

4 years agomodbus: Make C->S and S->C configurable, add framegap option.
Andrew Gregory [Wed, 5 Dec 2018 15:31:34 +0000 (23:31 +0800)]
modbus: Make C->S and S->C configurable, add framegap option.

Change client->server and server->client to be separately configurable,
allowing decoding at both the server (where client->server is RX and
server->client is TX) and client (where client->server is TX and
server->client is RX) ends of the link. It also allows monitoring of the
bus on a single channel (where client->server and server->client are both
RX (or TX)).

When I tried to decode a bus capture, I found that when the transmitter was
turned off it generated a false start bit, which in turn resulted in a false
trailing byte from the UART decoder. This narrowed the inter-frame gap to
the point where the Modbus decoder failed to recognise a new frame. The
result was only the first frame of the capture decoded - all the rest of the
frames failed to decode. I had to reduce the frame gap to allow subsequent
frames to decode, and so made it a configurable option that defaults to the
existing gap.

Lastly, I fixed a call to puti() that incorrectly included the annotation
prefix.

4 years agosdcard_spi: Add "Card is busy" annotations for CMD24.
Pierre Poulain [Tue, 16 Apr 2019 17:56:54 +0000 (19:56 +0200)]
sdcard_spi: Add "Card is busy" annotations for CMD24.

This fixes bug #1376.

4 years agosdcard_spi: Fix incorrect handling of CMD17.
Pierre Poulain [Tue, 16 Apr 2019 18:48:11 +0000 (20:48 +0200)]
sdcard_spi: Fix incorrect handling of CMD17.

This fixes bug #1377.

4 years agousb_request: Allow configuration of BULK IN transfer display
Stefan Brüns [Mon, 17 Jun 2019 18:01:58 +0000 (20:01 +0200)]
usb_request: Allow configuration of BULK IN transfer display

From a protocol level, a BULK IN transfer starts when the host starts
polling the respective endpoint. For analysis, it is sometimes useful
to show when the devices starts to answer the requests.

As both are useful for different use cases (the old, default one emphasizes
the host behavior, the new one shows the endpoint/device behavior), make
the display configurable.

4 years agousb_request: Use separate annotation rows for IN, OUT and CONTROL
Stefan Brüns [Mon, 17 Jun 2019 17:56:51 +0000 (19:56 +0200)]
usb_request: Use separate annotation rows for IN, OUT and CONTROL

E.g. CONTROL and BULK IN transfers may overlap each other, and as a result
only one of the two can be seen in pulseview.

Partly solves bug #1046. In case a device has multiple IN/OUT endpoints,
transfers would still overlap, but many simple devices have just one each.

4 years agousb_request: Fix some decoding errors
Stefan Brüns [Mon, 17 Jun 2019 17:51:29 +0000 (19:51 +0200)]
usb_request: Fix some decoding errors

Append OUT data only if it has been ACKed. OUT transfers (BULK OUT or
CONTROL transfer DATA stage) are typically NACKed to create backpressure.

Always keep IN and OUT transfers separate. On the physical layer, the
endpoint number only uses 4 bits, and IN and OUT use separate tokens.
In case the transfer is an IN transfer, set the high bit as used in the
endpoint descriptors (i.e. 0x81 is IN enpoint 1, 0x01 is OUT endpoint 1).

4 years agoenc28j60: Rename self.ann to self.out_ann.
Uwe Hermann [Tue, 14 May 2019 23:28:53 +0000 (01:28 +0200)]
enc28j60: Rename self.ann to self.out_ann.

(for consistency with all other decoders)

4 years agoenc28j60: Shorten various self.put*() calls.
Uwe Hermann [Tue, 14 May 2019 23:26:02 +0000 (01:26 +0200)]
enc28j60: Shorten various self.put*() calls.

4 years agoenc28j60: Factor out self.putr().
Uwe Hermann [Tue, 14 May 2019 23:21:24 +0000 (01:21 +0200)]
enc28j60: Factor out self.putr().

4 years agoenc28j60: Factor out self.putc().
Uwe Hermann [Tue, 14 May 2019 23:16:52 +0000 (01:16 +0200)]
enc28j60: Factor out self.putc().

4 years agoenc28j60: Use the common ss/es abbreviations.
Uwe Hermann [Tue, 14 May 2019 23:08:16 +0000 (01:08 +0200)]
enc28j60: Use the common ss/es abbreviations.

This makes it more consistent with the rest of the decoders.

4 years agoenc28j60: Clear 'outputs' (no OUTPUT_PYTHON support).
Uwe Hermann [Tue, 14 May 2019 22:59:44 +0000 (00:59 +0200)]
enc28j60: Clear 'outputs' (no OUTPUT_PYTHON support).

4 years agoenc28j60: Factor out command list to lists.py.
Uwe Hermann [Tue, 14 May 2019 22:57:42 +0000 (00:57 +0200)]
enc28j60: Factor out command list to lists.py.

4 years agoenc28j60: Shorten ANN_* list a bit.
Uwe Hermann [Tue, 14 May 2019 22:56:46 +0000 (00:56 +0200)]
enc28j60: Shorten ANN_* list a bit.

4 years agoenc28j60: Add 'tags' field.
Uwe Hermann [Tue, 14 May 2019 22:49:04 +0000 (00:49 +0200)]
enc28j60: Add 'tags' field.

4 years agoenc28j60: Implements decoder for ENC28J60 SPI Ethernet chip
Jiahao Li [Wed, 20 Feb 2019 22:38:11 +0000 (17:38 -0500)]
enc28j60: Implements decoder for ENC28J60 SPI Ethernet chip

4 years agoatsha204a: Mention other potentially supported devices.
Uwe Hermann [Tue, 30 Apr 2019 20:26:59 +0000 (22:26 +0200)]
atsha204a: Mention other potentially supported devices.

4 years agoatsha204a: Mention ATECC508A support in docs.
Uwe Hermann [Tue, 30 Apr 2019 20:15:54 +0000 (22:15 +0200)]
atsha204a: Mention ATECC508A support in docs.

4 years agoatsha204a: Minor indentation fixes.
Uwe Hermann [Tue, 30 Apr 2019 20:03:27 +0000 (22:03 +0200)]
atsha204a: Minor indentation fixes.

4 years agoatsha204a: Only call output_tx_bytes() when the bytes buffer is not empty
Michalis Pappas [Sat, 26 Jan 2019 21:28:04 +0000 (22:28 +0100)]
atsha204a: Only call output_tx_bytes() when the bytes buffer is not empty

4 years agoatsha204a: Display OtherData parameter correctly when Verify command is executed...
Michalis Pappas [Wed, 11 Jul 2018 17:45:15 +0000 (20:45 +0300)]
atsha204a: Display OtherData parameter correctly when Verify command is executed in Validate mode.

4 years agoatsha204a: Add support for ATECC508A
Michalis Pappas [Fri, 15 Jun 2018 09:29:21 +0000 (12:29 +0300)]
atsha204a: Add support for ATECC508A

4 years agoNEWS: Fix two typos.
Uwe Hermann [Wed, 24 Apr 2019 20:02:59 +0000 (22:02 +0200)]
NEWS: Fix two typos.

4 years agospi: Fix annotation row ordering.
Uwe Hermann [Wed, 24 Apr 2019 20:00:02 +0000 (22:00 +0200)]
spi: Fix annotation row ordering.

Use the same (more logical) order of having the "bits" annotation row
being the first one, which is also what pretty much all other PDs do.

4 years agospi: Don't emit bogus "transfer" items.
Uwe Hermann [Wed, 24 Apr 2019 19:57:35 +0000 (21:57 +0200)]
spi: Don't emit bogus "transfer" items.

4 years agospi: Add support for "transfer" annotations.
Uwe Hermann [Wed, 24 Apr 2019 19:10:40 +0000 (21:10 +0200)]
spi: Add support for "transfer" annotations.

There is already an OUTPUT_PYTHON type named 'TRANSFER' that is being
transmitted to stacked decoders. Make the same information available
as annotations (additionally), which can be helpful in some situations.

5 years agocc1101: Use namedtuple to increase readability a bit.
Uwe Hermann [Fri, 19 Apr 2019 13:42:15 +0000 (15:42 +0200)]
cc1101: Use namedtuple to increase readability a bit.

5 years agocc1101: Remove 0x prefixes and unneeded quotation marks.
Uwe Hermann [Thu, 18 Apr 2019 22:25:31 +0000 (00:25 +0200)]
cc1101: Remove 0x prefixes and unneeded quotation marks.

The format "0xAA BB CC ..." looks a bit strange, use "AA BB CC" instead,
it's relatively obvious that the data is hex-encoded.

The quotation marks around some names/values are unneeded and reduce
readability a bit, so drop them.

5 years agocc1101: Shorten decode_register() to decode_reg().
Uwe Hermann [Thu, 18 Apr 2019 22:24:28 +0000 (00:24 +0200)]
cc1101: Shorten decode_register() to decode_reg().

5 years agocc1101: Simplify decode_command().
Uwe Hermann [Thu, 18 Apr 2019 22:20:43 +0000 (00:20 +0200)]
cc1101: Simplify decode_command().

By using the annotation strings as command names directly, some parts
of decode_command() can be simplified a bit.

5 years agocc1101: Simplify format_command().
Uwe Hermann [Thu, 18 Apr 2019 22:02:54 +0000 (00:02 +0200)]
cc1101: Simplify format_command().

5 years agocc1101: Factor out and simplify ANN_* names.
Uwe Hermann [Thu, 18 Apr 2019 21:55:39 +0000 (23:55 +0200)]
cc1101: Factor out and simplify ANN_* names.

5 years agoAdd a TI CC1101 decoder.
Marco [Wed, 27 Mar 2019 22:47:36 +0000 (23:47 +0100)]
Add a TI CC1101 decoder.

5 years agosdcard_spi: Fix duplicated 'End bit' annotation emission.
Uwe Hermann [Mon, 15 Apr 2019 18:19:19 +0000 (20:19 +0200)]
sdcard_spi: Fix duplicated 'End bit' annotation emission.

5 years agosdcard_spi: Fix incorrect annotation row indices.
Uwe Hermann [Mon, 15 Apr 2019 18:13:32 +0000 (20:13 +0200)]
sdcard_spi: Fix incorrect annotation row indices.

This fixes the remaining parts of bug #1365.

5 years agosdcard_spi: handle_data_response(): Shorten a few lines.
Uwe Hermann [Mon, 15 Apr 2019 17:46:04 +0000 (19:46 +0200)]
sdcard_spi: handle_data_response(): Shorten a few lines.

5 years agosdcard_spi: handle_data_response(): Add comment, cosmetics.
Uwe Hermann [Mon, 15 Apr 2019 17:40:55 +0000 (19:40 +0200)]
sdcard_spi: handle_data_response(): Add comment, cosmetics.

5 years agosdcard_spi: Fix some abbreviations and names to match the spec.
Uwe Hermann [Mon, 15 Apr 2019 16:50:12 +0000 (18:50 +0200)]
sdcard_spi: Fix some abbreviations and names to match the spec.

The official abbreviation for CMD24 in the spec is "WRITE_BLOCK", as
opposed to "WRITE_MULTIPLE_BLOCK" for CMD25 (chapter 4.7.4,
"Detailed Command Description", table 4-24).

The byte preceeding e.g. the CMD24 block data is called "Start Block" token
(chapter 7.3.3.2, "Start Block Tokens and Stop Tran Token"). We don't
include the "token" itself for consistency, since the decoder doesn't do
that for any other tokens either.

5 years agosdcard_spi: touch up CMD24 handling
Gerhard Sittig [Wed, 3 Apr 2019 17:10:46 +0000 (19:10 +0200)]
sdcard_spi: touch up CMD24 handling

Address some nits in the SDCard (SPI mode) protocol decoder. Rename
identifiers to eliminate comments. Determine the default block size at
the start of the write command instead of the iteration over payload
data bytes. Remove a print() statement which would break regression
tests. Allow re-use of the data handler for other commands, too. Use
lower case hex digits for consistency across the source file, and
slightly unobfuscate a bit pattern check while we are here. Improve
robustness of response handlers and how internal state gets advanced.
Replace constant lookups by direct method calls.

5 years agosdcard_spi: add support for CMD24 including payload data
Pierre Poulain [Fri, 15 Mar 2019 17:37:24 +0000 (18:37 +0100)]
sdcard_spi: add support for CMD24 including payload data

The previous implementation advanced from the start of CMD24 to IDLE.
This commit introduces support for the data phase and data response of
CMD24. Which results in improved usability of the decoder output, and
reduced probability of false positives (don't detect "commands" in the
payload data).

This commit addresses the missing CMD24 handling part of bug #1365.

5 years agosrd_inst_stack(): Warn upon potentially incorrect stacking.
Uwe Hermann [Tue, 2 Apr 2019 19:35:32 +0000 (21:35 +0200)]
srd_inst_stack(): Warn upon potentially incorrect stacking.

Currently there must be at least one match of inputs/outputs for
decoders that are stacked. If not, we emit an informational warning
(but this is not a hard error for the time being).

5 years agodecoders: Fix incorrect 'outputs' fields.
Uwe Hermann [Tue, 2 Apr 2019 19:33:58 +0000 (21:33 +0200)]
decoders: Fix incorrect 'outputs' fields.

Only add items to 'outputs' if the respective PD actually has
OUTPUT_PYTHON support implemented as of right now.

Various decoders might get OUTPUT_PYTHON support later, but the
'outputs' field should reflect the current status.

5 years agods2408: Add missing tags.
Uwe Hermann [Thu, 28 Mar 2019 21:49:39 +0000 (22:49 +0100)]
ds2408: Add missing tags.

5 years agoadd a Maxim DS2408 decoder
Mariusz Bialonczyk [Thu, 14 Mar 2019 11:25:54 +0000 (12:25 +0100)]
add a Maxim DS2408 decoder

This is a 1-Wire 8-channel addressable switch.

5 years agoonewire_network: add a 'Disable Test Mode' command for DS2408
Mariusz Bialonczyk [Sat, 16 Mar 2019 10:57:22 +0000 (11:57 +0100)]
onewire_network: add a 'Disable Test Mode' command for DS2408

For details see the DS2408 datasheet, section "Power-up timing".

This commit is adding this missing command so now a full decoding
is possible:
Annotations from 1-Wire network layer:
8990980-8991220 1-Wire network layer: : Reset/presence: true
8994163-8994437 1-Wire network layer: : ROM command: 0x96 'DS2408: Disable Test Mode'
8995026-9008178 1-Wire network layer: : ROM: 0xa30000001246a929
9008785-9009060 1-Wire network layer: : Data: 0x3c
9012290-9012530 1-Wire network layer: : Reset/presence: true

And from the specific stacked DS2408 slave:
8990980-8991220 DS2408: : Reset/presence: true
8995026-9008178 DS2408: : ROM: 0xa30000001246a929 (family code 0x29)
9008785-9009060 DS2408: : Disable Test Mode (0x3c)
9012290-9012530 DS2408: : Reset/presence: true

5 years agoonewire_network: add a missing 'Resume ROM' command
Mariusz Bialonczyk [Wed, 13 Mar 2019 17:47:38 +0000 (18:47 +0100)]
onewire_network: add a missing 'Resume ROM' command

For details see the DS2408 datasheet, section "Resume Command [A5h]".

5 years agodecoders: Various cosmetic/consistency/typo fixes.
Uwe Hermann [Sat, 16 Mar 2019 17:46:01 +0000 (18:46 +0100)]
decoders: Various cosmetic/consistency/typo fixes.

5 years agoHACKING: Document current tags and their intention/usage.
Uwe Hermann [Thu, 14 Mar 2019 23:57:31 +0000 (00:57 +0100)]
HACKING: Document current tags and their intention/usage.

5 years agodecoders: Add/update tags for each PD.
Uwe Hermann [Thu, 14 Mar 2019 23:57:04 +0000 (00:57 +0100)]
decoders: Add/update tags for each PD.

5 years agoAdd PD tags handling and some tags
Soeren Apel [Fri, 28 Dec 2018 10:41:41 +0000 (11:41 +0100)]
Add PD tags handling and some tags

5 years agojtag_ejtag: Shorten a variable name.
Uwe Hermann [Tue, 29 Jan 2019 23:03:23 +0000 (00:03 +0100)]
jtag_ejtag: Shorten a variable name.

5 years agoRename values that shadow built-in keywords
Ryan Jarvis [Sat, 19 Jan 2019 08:05:07 +0000 (00:05 -0800)]
Rename values that shadow built-in keywords

5 years agoRemove trailing semicolons
Ryan Jarvis [Sat, 19 Jan 2019 07:51:23 +0000 (23:51 -0800)]
Remove trailing semicolons

5 years agoFix obvious typos with string formatters
Ryan Jarvis [Sat, 19 Jan 2019 07:49:47 +0000 (23:49 -0800)]
Fix obvious typos with string formatters

5 years agoTypo preventing "Personal preference settings" & "Display user info on screen" from...
Ryan Jarvis [Sat, 19 Jan 2019 07:46:36 +0000 (23:46 -0800)]
Typo preventing "Personal preference settings" & "Display user info on screen" from functioning

Source: https://en.wikipedia.org/wiki/RC-5#Command_Tables

5 years agoComparison to None should use "is/is not" instead of "==/!="
Ryan Jarvis [Sat, 19 Jan 2019 07:43:26 +0000 (23:43 -0800)]
Comparison to None should use "is/is not" instead of "==/!="

5 years agoDS1307: Bugs fixes.
Libor Gabaj [Wed, 23 Jan 2019 20:58:26 +0000 (21:58 +0100)]
DS1307: Bugs fixes.

- Square wave frequencies above 1 Hz are in Hz not in kHz.
- AM/PM flag is in the bit 5 of hours register not in bit 6.
- AM flag is valid at 0 value of AM/PM flag not at 1 value.

5 years agouart: remove obsolete TODO (Python annotation for frame errors)
Gerhard Sittig [Sat, 24 Nov 2018 10:57:38 +0000 (11:57 +0100)]
uart: remove obsolete TODO (Python annotation for frame errors)

The "Frame error?" TODO comment on Python annotations has become
obsolete. Individual bit errors within the frame immediately get
communicated as they are detected (START, parity, STOP). The overall
frame's validity has become available with the FRAME annotation.

5 years agouart: add FRAME Python annotation, communicate frame validity
Gerhard Sittig [Fri, 23 Nov 2018 18:10:34 +0000 (19:10 +0100)]
uart: add FRAME Python annotation, communicate frame validity

Internally keep track of the UART frame's validity. Emit a FRAME Python
annotation for aborted as well as for completed frames. This obsoletes a
TODO comment in the STOP bit code path.

This annotation also spans the complete frame's length, including start
and parity and stop bits, which the DATA annotation doesn't cover.
Stacked decoders can individually decide whether to strictly reference
the mere data bits section or the complete UART frame which happened to
communicate the data value.

5 years agouart: document the Python annotation for BREAK
Gerhard Sittig [Fri, 23 Nov 2018 18:09:09 +0000 (19:09 +0100)]
uart: document the Python annotation for BREAK

5 years agoi2s: Fix 2 issues.
Weixiao Zhang [Wed, 21 Nov 2018 03:41:45 +0000 (21:41 -0600)]
i2s: Fix 2 issues.

1. WS = 0, Left Channel; WS = 1, Right Channel
2. Data start and end at falling edge of SCK

5 years agox2444m: Add some shorter annotation strings as well.
Uwe Hermann [Sat, 10 Nov 2018 17:48:37 +0000 (18:48 +0100)]
x2444m: Add some shorter annotation strings as well.

5 years agox2444m: Move variable initializations to reset().
Uwe Hermann [Sat, 10 Nov 2018 17:31:38 +0000 (18:31 +0100)]
x2444m: Move variable initializations to reset().

5 years agox2444m: Make each command an extra annotation class.
Uwe Hermann [Sat, 10 Nov 2018 17:31:04 +0000 (18:31 +0100)]
x2444m: Make each command an extra annotation class.

5 years agoDecoder for Xicor X2444M, nonvolatile static RAM.
Stefan Petersen [Wed, 31 Oct 2018 13:27:44 +0000 (14:27 +0100)]
Decoder for Xicor X2444M, nonvolatile static RAM.

5 years agodecoder: consider reset() a mandatory decoder method
Gerhard Sittig [Sun, 21 Oct 2018 15:14:59 +0000 (17:14 +0200)]
decoder: consider reset() a mandatory decoder method

In theory support for PD reset is optional, applications are not
required to make use of it. But it's essential to receive correct
decoding results when used with the popular Pulseview mainline
application.

So let's consider the absence of the reset() method fatal. All mainline
decoders have it. Out-of-tree decoders are easy to adjust, and very
probably should support reset() as well. Thus this change is considered
beneficial, and not harmful.

5 years agomiller: add reset() method
Gerhard Sittig [Sun, 21 Oct 2018 19:25:34 +0000 (21:25 +0200)]
miller: add reset() method

5 years agolin: Simplify state handling a bit.
Uwe Hermann [Sun, 21 Oct 2018 17:43:10 +0000 (19:43 +0200)]
lin: Simplify state handling a bit.

5 years agolin: handle_break(): Use self.putx().
Uwe Hermann [Sun, 21 Oct 2018 17:30:32 +0000 (19:30 +0200)]
lin: handle_break(): Use self.putx().

5 years agolin: Simplify LinFsm._transition_allowed().
Uwe Hermann [Sun, 21 Oct 2018 17:28:31 +0000 (19:28 +0200)]
lin: Simplify LinFsm._transition_allowed().

5 years agoadd LIN protocol decoder
stephan.thiele [Wed, 17 Oct 2018 09:40:23 +0000 (11:40 +0200)]
add LIN protocol decoder

5 years agospiflash: Remember write protection latch from status registers
Vesa-Pekka Palmu [Wed, 17 Oct 2018 22:57:28 +0000 (01:57 +0300)]
spiflash: Remember write protection latch from status registers

5 years agospiflash: Bugfix: WRSR was using miso for register decode
Vesa-Pekka Palmu [Wed, 17 Oct 2018 22:56:26 +0000 (01:56 +0300)]
spiflash: Bugfix: WRSR was using miso for register decode

5 years agospiflash: Initial support for write enable state tracking
Vesa-Pekka Palmu [Wed, 17 Oct 2018 15:49:29 +0000 (18:49 +0300)]
spiflash: Initial support for write enable state tracking

5 years agospiflash: Handle chip erase
Vesa-Pekka Palmu [Wed, 17 Oct 2018 12:53:35 +0000 (15:53 +0300)]
spiflash: Handle chip erase

Minimal implementation of chip erase 0x60 and 0xc7 command handling

5 years agospiflash: Add Winbond W25Q80DV metadata.
Uwe Hermann [Sun, 21 Oct 2018 15:32:57 +0000 (17:32 +0200)]
spiflash: Add Winbond W25Q80DV metadata.

5 years agoFix GVariant memory leaks in session test code
Jon Burgess [Sun, 14 Oct 2018 18:11:21 +0000 (19:11 +0100)]
Fix GVariant memory leaks in session test code

==187759== 88 (40 direct, 48 indirect) bytes in 1 blocks are definitely lost in loss record 2,262 of 3,218
==187759==    at 0x4C2EBAB: malloc (vg_replace_malloc.c:299)
==187759==    by 0x563C435: g_malloc (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187759==    by 0x5654056: g_slice_alloc (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187759==    by 0x5673630: g_variant_new_from_bytes (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187759==    by 0x566C5E6: ??? (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187759==    by 0x566C7D2: g_variant_new_uint64 (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187759==    by 0x403D41: conf_check_fail.constprop.6 (session.c:152)
==187759==    by 0x403E06: test_session_metadata_set_bogus (session.c:215)
==187759==    by 0x53E51D5: srunner_run_tagged (in /usr/lib64/libcheck.so.0.0.0)
==187759==    by 0x401237: main (main.c:51)

5 years agoFix memory leak in decoder test code
Jon Burgess [Sun, 14 Oct 2018 18:10:16 +0000 (19:10 +0100)]
Fix memory leak in decoder test code

==187724== 917 bytes in 1 blocks are definitely lost in loss record 3,046 of 3,207
==187724==    at 0x4C2EBAB: malloc (vg_replace_malloc.c:299)
==187724==    by 0x563C435: g_malloc (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187724==    by 0x5655D82: g_strdup (in /usr/lib64/libglib-2.0.so.0.5600.3)
==187724==    by 0x4E43937: py_str_as_str (util.c:371)
==187724==    by 0x4E415E1: srd_decoder_doc_get (decoder.c:892)
==187724==    by 0x40281D: test_doc_get (decoder.c:395)
==187724==    by 0x53E51D5: srunner_run_tagged (in /usr/lib64/libcheck.so.0.0.0)
==187724==    by 0x401237: main (main.c:51)

5 years agoFix memory leak in print_searchpaths()
Jon Burgess [Sun, 14 Oct 2018 17:55:47 +0000 (18:55 +0100)]
Fix memory leak in print_searchpaths()

==175453== 522 bytes in 8 blocks are definitely lost in loss record 2,923 of 3,201
==175453==    at 0x4C2EBAB: malloc (vg_replace_malloc.c:299)
==175453==    by 0x59E9BB5: PyObject_Malloc (in /usr/lib64/libpython3.6m.so.1.0)
==175453==    by 0x5A35A76: PyBytes_FromStringAndSize (in /usr/lib64/libpython3.6m.so.1.0)
==175453==    by 0x4E3FA6E: print_searchpaths (srd.c:173)
==175453==    by 0x4E3FA6E: srd_init (srd.c:287)
==175453==    by 0x4034BE: test_session_reset_nodata (session.c:238)
==175453==    by 0x53E51D5: srunner_run_tagged (in /usr/lib64/libcheck.so.0.0.0)
==175453==    by 0x401237: main (main.c:51)

5 years agoFix bad memory accesses during srd_exit()
Jon Burgess [Sun, 14 Oct 2018 17:49:52 +0000 (18:49 +0100)]
Fix bad memory accesses during srd_exit()

When the decoder and session unload functions are called they
remove themselves from the list. The code walking the list
must be careful to avoid accessing the next pointer which
might now be invalid. The g_slist_foreach() takes care of
this.

Reports from Valgrind before fix:

==175436== Invalid read of size 8
--
==175436==  Address 0xe3f2598 is 8 bytes inside a block of size 16 free'd
==175436==    at 0x4C2FDAC: free (vg_replace_malloc.c:530)
==175436==    by 0x563C541: g_free (in /usr/lib64/libglib-2.0.so.0.5600.3)
==175436==    by 0x5654783: g_slice_free1 (in /usr/lib64/libglib-2.0.so.0.5600.3)
==175436==    by 0x56552A2: g_slist_remove (in /usr/lib64/libglib-2.0.so.0.5600.3)
==175436==    by 0x4E3FEFF: srd_session_destroy (session.c:343)
==175436==    by 0x4E3F5C7: srd_exit (srd.c:311)
==175436==    by 0x40336F: test_inst_new (inst.c:40)
==175436==    by 0x53E51D5: srunner_run_tagged (in /usr/lib64/libcheck.so.0.0.0)
==175436==    by 0x401237: main (main.c:51)
==175436==  Block was alloc'd at
==175436==    at 0x4C2EBAB: malloc (vg_replace_malloc.c:299)
==175436==    by 0x563C435: g_malloc (in /usr/lib64/libglib-2.0.so.0.5600.3)
==175436==    by 0x5654056: g_slice_alloc (in /usr/lib64/libglib-2.0.so.0.5600.3)
==175436==    by 0x5655797: g_slist_append (in /usr/lib64/libglib-2.0.so.0.5600.3)
==175436==    by 0x4E3FC75: srd_session_new (session.c:71)
==175436==    by 0x403345: test_inst_new (inst.c:37)
==175436==    by 0x53E51D5: srunner_run_tagged (in /usr/lib64/libcheck.so.0.0.0)
==175436==    by 0x401237: main (main.c:51)

5 years agoUpdate a few Doxygen @since tags.
Uwe Hermann [Sun, 14 Oct 2018 22:16:51 +0000 (00:16 +0200)]
Update a few Doxygen @since tags.

Drop @since tags for non-public functions.

5 years agoinstance.c: Avoid 4 unneeded forward declarations.
Uwe Hermann [Sun, 14 Oct 2018 22:09:41 +0000 (00:09 +0200)]
instance.c: Avoid 4 unneeded forward declarations.

5 years agouart: add support for break condition detection
Gerhard Sittig [Sun, 14 Oct 2018 18:17:35 +0000 (20:17 +0200)]
uart: add support for break condition detection

There are the "traffic inspecting" wait() conditions, which check an
edge to find the start of START, then wait for sample points to grab the
bit values. Bit times are sampled in their respective center, potential
glitches around sample points get ignored.

Add another independent set of wait() conditions which check _all_ edges
regardless of any data communication. This results in the most reliable
and maintainable detection of break conditions, regardless of how they
align to data frames. Break is defined as a period of low input signal
which spans at least one frame's length. Run the edge inspection after
data inspection, which results in the most appropriate annotation output
like leading data bits (of incomplete frames), frame errors (violated
STOP bit expectations), then break conditions. This approach is most
robust in the presence of incomplete input streams.

5 years agoook: Fix two typos which affect the decoding results.
Uwe Hermann [Sat, 13 Oct 2018 15:49:04 +0000 (17:49 +0200)]
ook: Fix two typos which affect the decoding results.

5 years agoVarious log message improvements.
Uwe Hermann [Fri, 12 Oct 2018 23:29:03 +0000 (01:29 +0200)]
Various log message improvements.

5 years agolog: Use human-readable output type name everywhere.
Uwe Hermann [Fri, 12 Oct 2018 23:02:30 +0000 (01:02 +0200)]
log: Use human-readable output type name everywhere.

5 years agoinstance.c: Make oldpins_array_*() static.
Uwe Hermann [Fri, 12 Oct 2018 22:53:22 +0000 (00:53 +0200)]
instance.c: Make oldpins_array_*() static.

5 years agoDecoder_put(): Add more info to OTUPUT_PYTHON log messages.
Uwe Hermann [Fri, 12 Oct 2018 22:49:06 +0000 (00:49 +0200)]
Decoder_put(): Add more info to OTUPUT_PYTHON log messages.

5 years agosrd.c: Fix a compiler warning with gcc 8.
Uwe Hermann [Sat, 6 Oct 2018 13:57:49 +0000 (15:57 +0200)]
srd.c: Fix a compiler warning with gcc 8.

  srd.c: In function ‘srd_searchpaths_get’:
  srd.c:399:40: warning: cast between incompatible function types from ‘gchar * (*)(const gchar *)’ {aka ‘char * (*)(const char *)’} to ‘void * (*)(const void *, void *)’ [-Wcast-function-type]
    return g_slist_copy_deep(searchpaths, (GCopyFunc)g_strdup, NULL);
                                          ^

Upstream glib issue / documentation change:
https://gitlab.gnome.org/GNOME/glib/issues/1492
https://gitlab.gnome.org/pwithnall/glib/commit/e81f4c2acea5ada6ae989426e462613f7c612cac

5 years agotype_decoder.c: Fix trailing whitespace.
Uwe Hermann [Sat, 6 Oct 2018 13:39:53 +0000 (15:39 +0200)]
type_decoder.c: Fix trailing whitespace.