1
0
mirror of https://bitbucket.org/librepilot/librepilot.git synced 2024-11-30 08:24:11 +01:00
Commit Graph

64 Commits

Author SHA1 Message Date
Alessio Morale
f767d64f0c OP-980 Remove the timeout logic as it causing troubles with windows
platform
2013-06-11 10:27:32 +02:00
Corvus Corax
6f59c49506 just uncrustification 2013-06-09 22:09:04 +02:00
Alessio Morale
87db5d30b0 OP-980 change how the semaphore is used on PIOS_USB_CheckAvailable to prevent deadlocks 2013-06-09 16:14:44 +02:00
Alessio Morale
ee7887c406 OP-980 Extended the memory barrier macros and fixed barrier kind in pios_usb_hid
+review OPReview-501
2013-06-06 10:26:41 +02:00
Alessio Morale
c340bfc294 OP-980 Insert locking while checking status in PIOS_USB_CheckAvailable
+review OPReview-501
2013-06-06 01:42:26 +02:00
Alessio Morale
5fa73624da OP-980 Added memory barrier to preserve sequence to ensure callback validation
+review OPReview-501
2013-06-06 01:39:45 +02:00
Alessio Morale
b4f8b02a42 Revert "some optimizations, use real rx and tx buffer size instead of max buffer leghts"
It does not work in Windows so need further investigation prior to undo this revert
This reverts commit 901db6f828.
+review OPReview-501
2013-06-05 19:01:01 +02:00
Alessio Morale
8fb81ae17b OP-980 uncrustify
+review OPReview-501
2013-06-05 01:02:37 +02:00
Alessio Morale
f1f48b0ec6 OP-980 Fix usb reconnection issue for Revolution target.
step 3: add a timeout on rx_active to prevent rx locks
2013-06-05 00:54:08 +02:00
Alessio Morale
901db6f828 some optimizations, use real rx and tx buffer size instead of max buffer leghts 2013-06-05 00:53:50 +02:00
Alessio Morale
ce2bce353f OP-980 Fix usb reconnection issue for Revolution target.
step 2: check for cable disconnection (vbus) and issue a SW reconnection
2013-06-05 00:04:26 +02:00
Alessio Morale
14b87383b9 OP-980 Fix usb reconnection issue for Revolution target.
step 1: reset tx/rx_active flags on disconnection/reconnection
2013-06-04 01:52:44 +02:00
Sexy Pony
900f643bbd Reformat source code with 'make uncrustify_all' run twice. NO CODE CHANGES 2013-05-19 17:37:30 +03:00
Alessio Morale
ab7af16be3 Merge branch 'amorale/OP-901_lib_udates-usb_fixes' into HEAD
Conflicts:
	flight/pios/stm32f10x/library.mk
	flight/targets/boards/revoproto/firmware/pios_board.c
2013-05-19 15:01:08 +02:00
Alessio Morale
37eabb0931 Pointed type is explicitly used with sizeof when dealing with struct pointers.
+review OPReview-474
2013-05-19 12:00:16 +02:00
Alessio Morale
2e458b5652 OP-901 USB OTG/Device Drivers are not targeted for f103 (they are targeted at f105) so move again to F4 library folder instead of common 2013-05-18 20:53:28 +02:00
Alessio Morale
c1f767b12d Merge remote-tracking branch 'origin/amorale/OP-954_logfs_settings_in_internal_flash' into next
Conflicts:
	flight/pios/common/pios_flashfs_logfs.c
2013-05-18 15:53:25 +02:00
Alessio Morale
976ad90211 OP-901 Updated F4 CMSIS startup files, Moved USB OTG and USB Device library to common 2013-05-18 14:03:03 +02:00
Alessio Morale
f0c01d0eff OP-901 Updated F4 StdPeriph Driver Library to V1.1.0 2013-05-18 13:57:45 +02:00
Alessio Morale
67e41c34fc Merge remote-tracking branch 'origin/next' into amorale/usb_fixes 2013-05-17 22:46:42 +02:00
Alessio Morale
2735f8a620 OP-901 Updated common CMSIS files 2013-05-17 22:41:34 +02:00
Alessio Morale
13ee42a42a OP-901 Fixed errors due to unused parameter 2013-05-17 22:39:16 +02:00
Richard Flay (Hyper)
a6aedee9fa Merge branch 'next' into brian/support_inverted_USB_vsense 2013-05-17 05:54:42 +09:30
Richard Flay (Hyper)
ae14c13195 Merge branch 'next' into hyper/OP-951_add--Wshadow-to-flight-CFLAGS
Conflicts:
	flight/modules/ManualControl/manualcontrol.c
2013-05-16 06:28:56 +09:30
Brian Webb
156109bb88 Added support for an active low USB vsense line. 2013-05-14 20:37:13 -07:00
Stacey Sheldon
280f537644 f4 usb: ensure that we NAK on unused OUT endpoints
The ST USB code will automatically receive on any
endpoint that is opened but not in the NAK state.
Make sure we set OUT endpoints to NAK initially.

It also happily writes via a NULL pointer in ep->xfer_buff
which writes to address 0x0000_0000.  Since address
0x0 is aliased onto the internal flash by the BOOT0/1 pins
and the internal flash is (normally) in the LOCKED state,
this write puts the internal flash into an errored state.

This errored state means that writes to internal flash
are no longer allowed and all further writes fail.
2013-05-14 23:50:17 +02:00
Richard Flay (Hyper)
ed68fbe68d OP-951: Adds -Wshadow to flight CFLAGS and fixes resulting compilation breakage.
+review OPReview
2013-05-14 07:01:45 +09:30
Alessio Morale
30b755b7aa OP-901 fixed errors due to unused parameter and applied fix against fifo empty bug on usb_dcd_int.c 2013-05-11 23:49:52 +02:00
Alessio Morale
95a3e99b7c OP-901 updated STM32 USB OTG Driver (updated configuration) 2013-05-11 23:12:14 +02:00
Alessio Morale
24fcea168c OP-901 Updated STM32F4 USB Device Library 2013-05-11 23:12:13 +02:00
Alessio Morale
ccc24d844a OP-901 updated STM32 USB OTG Driver 2013-05-11 23:10:47 +02:00
lilvinz
1016b08b70 f4 usb: removed DCD_EP_Flush on endpoint TX
This has probably caused the problem that prevented tx after
a bus reconnection.
2013-05-11 22:41:05 +02:00
Stacey Sheldon
0e0571149b f4 usb: ensure that we NAK on unused OUT endpoints
The ST USB code will automatically receive on any
endpoint that is opened but not in the NAK state.
Make sure we set OUT endpoints to NAK initially.

It also happily writes via a NULL pointer in ep->xfer_buff
which writes to address 0x0000_0000.  Since address
0x0 is aliased onto the internal flash by the BOOT0/1 pins
and the internal flash is (normally) in the LOCKED state,
this write puts the internal flash into an errored state.

This errored state means that writes to internal flash
are no longer allowed and all further writes fail.
2013-05-11 22:41:03 +02:00
Stacey Sheldon
2165a5afe1 f4 usb hid: handle SET_IDLE request
The code doesn't do anything useful with the SET_IDLE
but this helps to keep the wireshark USB traces clean
so that the real errors stand out better.
2013-05-11 22:41:02 +02:00
Stacey Sheldon
c22daec1b0 usb cdc: avoid need for ZLP by using tx buffers smaller than max
USB CDC uses BULK endpoints to send/receive data.  Typically,
a USB host will enqueue large buffers on its IN (device-to-host)
URBs.  These buffers are larger than the max packet size for the
bulk endpoint.

The USB standard requires that an IN transfer ends when one of
these is true:
 * a short packet (ie. less than max packet size) is sent by the
   device
 * a zero length packet (ZLP)
 * enough packets that the entire host buffer is filled

Our device implementation never sends ZLPs.  We sometimes send
packets that are exactly max-packet-size bytes long.  This would
result in partially filling a host buffer without signalling (via
ZLP) that the transmission had finished.  The host would then wait
until the next transfer had taken place before processing the first
data, thus delaying the first data.

This change simply forces all of our transfers to be short packets
and avoids the need to worry about zero length packets.  This is
at the cost of some efficiency on the host side since its large
buffers will only ever be partially filled.

Conflicts:
	flight/PiOS/STM32F30x/pios_usb_cdc.c
2013-05-11 22:41:01 +02:00
Stacey Sheldon
366c573bab f4 usb hid: send correct report ID in dummy report
The desired report ID is in the LSB of the wValue field,
not the MSB.  This will now send the correct report ID
back to the host.
2013-05-11 22:41:01 +02:00
Stacey Sheldon
232f9b2f8f f4 usb hid: use static data for dummy report Tx
This code was previously passing a pointer to stack
data into PIOS_USBHOOK_CtrlTx() which may be sending
this data asynchronously.  Now pass a pointer to
static data so that the asynchronous send doesn't
tx random stack contents.
2013-05-11 22:41:00 +02:00
Stacey Sheldon
859734af98 f4 usb: fix tracking of OUT requests with data stages
This code was mistakenly tracking the IN (device-to-host)
requests with data stages.  It should have been tracking
the OUT (host-to-device) requests with data stages.

This tracking data is important now that CDC is supported.
CDC actually uses OUT requests with data stages whereas HID
does not.  This bug only triggered once CDC was enabled.
2013-05-11 22:41:00 +02:00
Stacey Sheldon
b56f5206bd com/usart/cdc: ensure that device structs are memset before use
CDC and USART device drivers were not all clearing their
device structs before using them.

This specifically caused crashes in the case where the upper
COM layer was binding only a Tx path.  The Rx path callback
in the lower driver was uninitialized random data and would
result in the lower driver faulting when it tried to call the
callback.

Conflicts:
	flight/PiOS/STM32F30x/pios_usart.c
	flight/PiOS/STM32F30x/pios_usb_cdc.c
	flight/PiOS/STM32F30x/pios_usb_hid.c
2013-05-11 22:40:59 +02:00
Stacey Sheldon
e16be98224 usb cdc: ensure com buffer is drained when no DTE is attached
The CDC layer on F1, F3 and F4 now always acts like an
infinte data sink whenever *either* there is no DTE present
(ie. no terminal program listening) *or* the USB cable is
disconnected.

F1 and F4 were previously checking the cable but not the DTE.
F3 didn't check anything.  The COM layer didn't even ask the
lower layers.

All of this used to mean that any time a caller did a blocking
send to a CDC device without a DTE, it would eventually block
for up to a 5s timeout waiting for space in the Tx buffer.

Conflicts:
	flight/PiOS/STM32F30x/pios_usb_cdc.c
2013-05-11 22:40:58 +02:00
Stacey Sheldon
c702ca0f1b usbhook: fix up indentation
No functional changes.
2013-05-11 22:40:57 +02:00
Stacey Sheldon
cb6f4a474e f4 usb: ensure data buffers are 4-byte aligned
4-byte alignment is required for DMA.
2013-05-11 22:40:56 +02:00
Stacey Sheldon
f4bed187f4 f4 usb cdc: implement CDC class driver for STM32 F4 SoCs 2013-05-11 22:38:26 +02:00
Stacey Sheldon
19714d2316 f4 usb: allocate TX FIFO RAM for endpoints 2 and 3
This is required to support USB CDC mgmt and data
endpoints.
2013-05-11 22:38:25 +02:00
Stacey Sheldon
02addaf4a0 usbhook: make ctrl endpoint OUT treat request as const
The req parameter is passed to the OUT stage as a convenience.
Since the OUT callback is called *after* the transfer has occurred,
modifying the req would have absolutely no effect.  Marking this
as const makes this expectation clear.
2013-05-11 22:38:25 +02:00
Alessio Morale
511283c608 Flash internal: writes optimization using 4 bytes parallelism 2013-05-11 22:04:47 +02:00
Richard Flay (Hyper)
04f59d5910 OP-950: Adds "higher priority task woken" logic to SPI and SPI bus resident
sensor device drivers. Based partly on code contributed by lilvinz.

+review OPReview
2013-05-11 16:18:34 +09:30
Corvus Corax
b76df471ee bugfix on revo proto. STM32F4 A revision has a hardware bug and cannot do flash prefetching properly 2013-05-10 19:53:51 +02:00
Richard Flay (Hyper)
cf14608ebe Merge branch 'next' into hyper/OP-936_task-monitor-rework 2013-05-09 19:25:02 +09:30
Alessio Morale
93bad2e8c2 Merge branch 'amorale/OP-917_f4_bootloader_updater' into next
***** PLEASE NOTE  ***** From this revision all F4 based boards (Revolution/OSD) needs bootloader version  5 to boot properly

Conflicts:
	flight/pios/stm32f4xx/libraries/CMSIS2/Device/ST/STM32F4xx/Source/osd/system_stm32f4xx.c
	flight/targets/boards/osd/board-info.mk
	make/boot-defs.mk
	make/common-defs.mk
2013-05-07 21:02:41 +02:00