f0b0b1bac7
support for multiple DDS buses (untested)
2016-03-09 17:12:50 +08:00
71105fd0d7
rtio: collision_error -> collision
2016-03-08 15:38:35 +08:00
whitequark
739568fcb8
runtime: fix sloppy memory management in cache_put.
2016-03-07 21:12:03 +00:00
125ab3e076
runtime/i2c_read: fix MSB
2016-03-05 19:00:14 +08:00
6b8efd10fd
runtime/i2c: fix artiq_raise_from_c invokation
2016-03-05 00:19:08 +08:00
ff4a46c278
runtime/i2c: make syscalls more ARTIQ-Python-friendly
2016-03-05 00:16:23 +08:00
c2fcefc31f
runtime/rtio: cleanup include
2016-03-03 19:48:06 +08:00
423ca03f3b
runtime: bit-banged i2c support (untested)
2016-03-03 17:46:42 +08:00
5ba753425d
runtime/rtio: rtio_process_exceptional_status() has only one user
2016-03-01 21:38:51 +01:00
f30dc4b39e
runtime: rt2wb_input -> rtio_input_data
2016-03-01 19:22:42 +01:00
81b35be574
bridge: really fix O/OE
2016-03-01 18:49:04 +01:00
135643e3a6
runtime: define constants for ttl addresses
2016-03-01 18:22:42 +01:00
6f9656dcbe
bridge: fix ttl o/oe addresses
2016-03-01 18:19:06 +01:00
8adef12781
runtime: refactor ttl*()
...
* remove rt2wb_output
* remove ttl_*() ttl.c ttl.h
* use rtio_output() and rtio_input_timestamp()
* adapt coredevice/compiler layer
* adapt bridge to not artiq_raise_from_c()
2016-03-01 16:36:59 +01:00
aa10791ddf
rtio: rm rtio_write_and_process_status
2016-03-01 15:40:35 +01:00
324660ab40
rt2wb, exceptions: remove RTIOTimeout
...
Assume that rt2wb transactions either collide and are then
reported (https://github.com/m-labs/artiq/issues/308 ) or that
they complete and the delay with which they complete does not matter.
If a transaction is ack'ed with a delay because the WB core's downstream
logic is busy, that may lead to a later collision with another WB
transaction.
2016-03-01 14:44:07 +01:00
7d7a710a56
runtime/rt2wb: use input/output terminology and add (async) input
2016-03-01 00:35:56 +01:00
7ab7f7d75d
Merge branch 'master' into spimaster
...
* master:
artiq_flash: use term 'gateware'
targets/kc705-nist_clock: add clock generator on LA32 for testing purposes
doc: insist that output() must be called on TTLInOut. Closes #297
doc: update install instructions
coredevice: do not give up on UTF-8 errors in log. Closes #300
use m-labs setup for defaults
fix indentation
2016-02-29 20:47:52 +01:00
6c899e6ba6
runtime/rtio: fix rtio_input_wait(), add RTIOTimeout
2016-02-29 19:49:15 +01:00
ecedbbef4c
runtime/ttl: use rtio_output and rtio_input_wait
2016-02-29 19:20:07 +01:00
5dae9f8aa8
runtime: refactor rt2wb/dds
2016-02-29 19:16:29 +01:00
d3c94827eb
runtime/ttl: simplify ttl_get() a bit
2016-02-29 17:58:54 +01:00
6903a1d88a
runtime/rt2wb: accurate exception strings
2016-02-29 14:56:04 +01:00
572c49f475
use m-labs setup for defaults
2016-02-29 21:35:23 +08:00
df7d15d1fe
runtime: refactor spi into rt2wb
2016-02-29 13:54:36 +01:00
aeae565d35
runtime/spi: don't apply channel offset
2016-02-29 11:53:36 +01:00
ad34927b0a
spi: RTIO_SPI_CHANNEL -> RTIO_FIRST_SPI_CHANNEL
2016-02-29 11:35:49 +01:00
d63a63531a
spi: add runtime support
2016-02-29 00:38:36 +01:00
whitequark
a1dd909bc4
Take alignment into account during attribute writeback ( fixes #293 ).
2016-02-25 01:44:05 +00:00
649f3b9094
Revert "Add workaround for analyzer core bug ( #206 )."
...
This reverts commit e94a9236aa
2016-02-15 09:57:23 +01:00
whitequark
e94a9236aa
Add workaround for analyzer core bug ( #206 ).
2016-02-15 03:56:56 +00:00
whitequark
74a75841f6
Add rtio_log() and make print() an RPC ( #206 ).
2016-02-15 03:56:56 +00:00
whitequark
b9448c069a
Make rtio_log() accept variable number of arguments.
2016-02-15 03:56:56 +00:00
whitequark
cd7f12e4b1
Remove lognonl(); rename log() to core_log() to avoid math.h conflict.
...
Also, make core_log() exactly equivalent to printf(), since the lack
of a drop-in universally usable printf() equivalent annoys me.
Note that this breaks compiled experiments. They'll be fixed in
the following commits.
2016-02-15 03:56:56 +00:00
10d78a5d0b
flash: grow runtime limit to 512 kB
2016-01-30 14:50:15 -07:00
e92d52314b
flash storage: move to flash + 0x70000
...
This requires recompiling and flashing the runtime and the entire
storage area file system (ip address and idle kernel).
* with ppp the runtime is larger than 0x40000 now
* 0x60000 for the runtime should give enough space for a while,
even including a fatter scheduler or runtime components
in other languages
* for both pipistrello and kc705 this leaves the unused space
starting at a nice round number
* the relevant flashes nowadays are 0x1000000 large
2016-01-28 18:08:30 -07:00
0079aa4037
runtime/main: style, wording, 1s ppp holdoff
2016-01-25 13:38:58 -07:00
2bc2cd0064
ppp: remove unneeded objects
2016-01-18 21:26:36 -07:00
1adeeabbed
ppp: wait for connection
2016-01-18 21:04:51 -07:00
dc709a77b8
PPP support (TCP broken)
2016-01-18 20:09:10 -07:00
whitequark
e2910fba99
runtime: define NDEBUG to disable debug mode of __artiq_personality.
...
There shouldn't be any bugs left and all the printing it does slows
exception handling considerably.
2016-01-19 01:52:57 +00:00
whitequark
5c6b1517d0
Rigorously treat builtin core device exceptions.
2016-01-19 01:45:25 +00:00
53b06a0b9b
RTIOUnderflow: show slack
2016-01-18 17:28:37 -07:00
15039e1d74
runtime/dds: DDS_ONEHOT_SEL -> CONFIG_DDS_ONEHOT_SEL
2016-01-15 09:32:17 -07:00
whitequark
1b9713115c
runtime: link libm and libbase-nofloat where appropriate.
2016-01-14 21:02:23 +00:00
whitequark
225f7d7302
Commit missing parts of 9366a29.
2016-01-10 20:01:26 +00:00
whitequark
d1319b83e5
runtime: properly format addresses in InternalError message.
2016-01-10 18:07:08 +00:00
whitequark
bbd0ca7827
Commit missing parts of 9366a29.
2016-01-10 15:45:56 +00:00
whitequark
f8e50f2b0c
Remove redundant ksupport API.
2016-01-10 14:27:46 +00:00
whitequark
cc45694f5a
Commit missing parts of 9366a29.
2016-01-10 13:08:26 +00:00
