Commit Graph

189 Commits

Author SHA1 Message Date
shaolin 8fd0a7f9c6 Update USB Device Stack to support more device controllers;
Pass USB CV test verification;
Code cleanup;
2014-02-16 17:42:35 +08:00
Bernard Xiong ec7fada97e [Drivers] Fix the block start issue in MTD Nand operators 2013-12-30 18:14:09 +08:00
bernard 42ae1c634d [Drivers] Add more debug utilities in MTD nand 2013-12-29 22:27:33 +08:00
Grissiom 6bb2c0f2ea Revert "serial: set the errno if and only if in thread context"
This reverts commit 07bd09fc0f.
The validation of the context is done with in `rt_set_errno`. So it's
safe to call it here.
2013-10-25 21:23:15 +08:00
Grissiom 07bd09fc0f serial: set the errno if and only if in thread context
The errno is associated with threads. It will crash the thread if we set
it in interrupt context.
2013-10-25 20:31:02 +08:00
Grissiom 21fb9a0bac serial: do onthing if the size of read/write is zero
Return early if the size is 0 as there is nothing to do.
2013-10-25 20:15:08 +08:00
Grissiom 0feefc03b6 device/pipe: add a control cmd to get the space left in pipe 2013-10-23 01:43:18 +08:00
joyhope 281e55e973 notification function: rt_err_t (*tx_complete)(rt_device_t dev, void
*buffer) with buffer pointer
2013-09-24 20:57:00 +08:00
Grissiom 246bdde8de dev/portal: fix bug in portal callback forwarding
When forwarding rx/tx callback from the underlaying device(pipe), the
"dev" argument should be the portal. So the portal callback could think
itself as called.
2013-08-22 16:43:06 +08:00
Grissiom 824261d3bc dev/portal: add checks on oflag
It does not make sense to open portal without a flag. One should open a
portal with RT_DEVICE_OFLAG_RDWR in most cases.
2013-08-22 15:12:00 +08:00
Grissiom 6e676e7754 dev/portal: implement portal device
Portal is a device that connect devices. Currently, you can only connect
pipes in portal. Pipes are unidirectional. But with portal, you can
construct a bidirectional device with two pipes.
2013-08-20 12:54:43 +08:00
Grissiom 7e68096a88 dev/pipe: make pipe a type of device
Pipe have many differences with char device. The main difference is
altough pipe have both read and write, it is unidirectional in nature.
2013-08-20 10:16:55 +08:00
Grissiom 0340300f89 dev/pipe: cleanup 2013-08-19 16:25:17 +08:00
Grissiom e8cca2315a dev/pipe: call rx_indicate the time we resume reader
Some reader(FinSh again) would like to block on a semaphore which is
released in rx_indicate. So we should invoke rx_indicate in
_rt_pipe_resume_reader.
2013-08-19 15:57:59 +08:00
Grissiom d683d32bd5 device/pipe: add nonblocking read/write and force write mode
The previous implementation will always blocks the reader/writer.
However, at least FinSh would expect the device to be nonblocking ---
read should return 0 when there is no data in it.
2013-08-19 15:35:56 +08:00
Grissiom 4919d29d69 ringbuffer: add put_force and putchar_force API
Add the APIs that will discard the old data when rb is full.
2013-08-19 15:32:42 +08:00
Grissiom 981d929b56 ringbuffer: fix the ambiguous name
RT_RINGBUFFER_SIZE could mean "the size of the whole buffer", "the size
of the empty space" or "the size of the data". Moreover, it's never a
micro anymore. Change it to rt_ringbuffer_data_len before it's too late.
Also, RT_RINGBUFFER_EMPTY is changed to rt_ringbuffer_space_len.
2013-08-19 14:12:35 +08:00
Grissiom 38ce3879d8 device/pipe: add rt_pipe_{init,detach}
This provide the possibility that allocate the buffer of the ringbuffer
on a specific region, instead of always mallocing it. It also bring us
the benefit of using pipe device on the systems without heap.
2013-08-19 12:28:59 +08:00
Bernard Xiong a431a25699 Merge pull request #135 from reynoldxu/K60Fxxxx
K60 fxxxx
2013-07-25 03:55:52 -07:00
reynoldxu 10c75c3dc6 remove a var for INT_TX. 2013-07-25 10:18:33 +08:00
reynoldxu 8f0fffc132 fix a bug in rt_hw_uart_init() where u CAN NOT init some feature. 2013-07-24 18:34:07 +08:00
bernard a124e7bd22 add initialization debug code 2013-07-24 07:33:48 +08:00
yiyue.fang 03cf76e7ef reformat the coding style in usbdevice/core/core.c 2013-06-29 10:32:27 +08:00
yiyue.fang 3359419d4f additional update 2013-06-28 00:39:11 +08:00
yiyue.fang 621fbdb90c Embedded GPLv2 license in drivers 2013-06-28 00:36:54 +08:00
Bernard Xiong 98c8a23e94 Add component initialization. 2013-06-23 07:48:42 +08:00
heyuanjie87 02869d7e67 udisk can be eject 2013-06-15 23:02:50 +08:00
Grissiom b02993c989 usbdevice/core: cope with some buggy drivers on USB_MSG_DATA_NOTIFY
Some driver will emit USB_MSG_DATA_NOTIFY before the USB get configured.
Filter out those events.
2013-06-11 11:50:05 +08:00
Grissiom ed19483cb4 usb: add USB_MSG_PLUG_OUT event
When the core received an USB_MSG_PLUG_OUT event, it will stop all the
classes. This make a chance that the classes could get rid off doing
useless stuff while the USB cable is plugged out.
2013-06-06 15:21:40 +08:00
heyuanjie87 376a08d89a return correct length of LangID string to host 2013-05-30 18:26:52 +08:00
weety 624d947e56 use __rt_ffs instead of ffs 2013-05-25 18:57:01 +08:00
qiuyiuestc a33e992724 Merge pull request #95 from grissiom/vcom
usbdevice/core: add reset function to support hotplug
2013-05-25 01:05:00 -07:00
qiuyiuestc 97a507e667 usbdevice/core: add reset function to support hotplug
When the USB got RESET packet from the host and the address is setup,
all the classes will got reset. The reset is done by class stop and than
class run. So the classes should reset their internal state in
class_{run,stop}.

Besides, the USB device driver could post a USB_MSG_RESET message on
every RESET packet.
2013-05-25 15:55:27 +08:00
qiuyiuestc 2e25159a83 Merge pull request #94 from grissiom/vcom
cdc_vcom: reset the vcom state in proper cases
2013-05-25 00:52:05 -07:00
Grissiom 046182be94 cdc_vcom: reset the vcom state in proper cases
Reset the vcom state in class_{run,stop} and when the PC is not
receiving data from me.
2013-05-25 15:25:32 +08:00
heyuanjie87 8aa4eb7924 make Host can halt endpoints 2013-05-23 12:33:22 +08:00
qiuyiuestc e822f87b1a Merge pull request #90 from heyuanjie87/ForPullRequest
For pull request
2013-05-17 06:39:05 -07:00
aozima c77c69c46e fix spelling error. 2013-05-17 20:59:38 +08:00
aozima 9a671a5002 move initial alarm event & mutex in system init. 2013-05-17 20:58:40 +08:00
aozima 36fc5520cb format code by Astyle. 2013-05-17 20:57:05 +08:00
heyuanjie87 8b4aadb4db add usb state 2013-05-17 20:53:05 +08:00
Grissiom b3e38f53b4 usbdevice/core: use static message queue
Although currently the only message send to the queue is struct
udev_msg, in order to keep backward compacity and for extention in the
future, we set the size of message to 32.
2013-05-12 15:43:30 +08:00
Grissiom f0d50a7b36 usbdevice/core: use static thread instead of dynamic thread
It also add two configurations for the USB device even loop thread:

    RT_USBD_THREAD_STACK_SZ to set the stack size, default to 2048
    RT_USBD_THREAD_PRIO     to set the priority, default to 8

You can overwrite the default values in rtconfig.h
2013-05-12 15:43:29 +08:00
Bernard Xiong f1e34eef11 Merge pull request #86 from grissiom/serial-rb-size
serial: use RT_SERIAL_RB_BUFSZ to define the size of ring buffer.
2013-05-11 23:18:18 -07:00
Bernard Xiong 8e9e7f3e7c serial: use RT_SERIAL_RB_BUFSZ to define the size of ring buffer. 2013-05-12 14:03:11 +08:00
Grissiom aa179e4438 cdc_vcom: not to start sending data when the data is already sending
We start the sending transaction in SOF handler. But if the data is
already sending, start an other transaction will cause data lose.
Implement a state machine is cdc_vcom and avoid that.
2013-05-12 13:27:13 +08:00
Grissiom 157af94af9 usbdevice/core: small optimization on the event loop
Sort the switch cases by the frequency of the events.
2013-05-12 13:07:03 +08:00
Grissiom d23ee75d2e cdc_vcom: send data as many as possible
The data filled into dcd_ep_write does not to be limited by
MaxPacketSize.
2013-05-10 17:50:07 +08:00
Grissiom 5de59429fe usbdevice/core: misc cleanup 2013-05-10 17:35:33 +08:00
Grissiom ffb1e62020 cdc_vcom: send a zero-length-packet at the end is transaction 2013-05-10 15:17:59 +08:00