andy/tinyUSB
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
8,999 Commits 2 Branches 0 Tags
c1d23a0a92bef016e22be3ed8194230cfab1c358
Commit Graph

4160 Commits

Author SHA1 Message Date
hathach
c1d23a0a92 osal_spin skipping lock/unlock when executed in isr 2025-05-21 11:19:07 +07:00
hathach
a4875fefea rename osal_critcal to osal_spinlock 
add spinlock implementation for most rtos
 2025-05-20 17:01:31 +07:00
hathach
bb1d348eb3 implement osal critical for none/freertos/pico-sdk 2025-05-19 22:51:40 +07:00
hathach
9aea7751f2 dwc2 only enter critical in isr with multiple core CPUs 2025-05-19 20:04:20 +07:00
hathach
72ee742761 add osal_critical API() for use with dwc2 2025-05-19 20:04:19 +07:00
Tomas Rezucha
c8baba10f9 fix(dcd): Fixed race condition on device disconnect 
TinyUSB does not provide any locking means to protect the DCD variables.
This can lead to race conditions when the user is trying to submit
a transfer while the device is being disconnected. This can cause
the device to be in an inconsistent state, leading to a crash or
undefined behavior.

This commit adds a spin-lock to protect the DCD variables during
device disconnect.

Closes https://github.com/espressif/esp-idf/issues/9691
Also reported in https://github.com/espressif/esp-usb/pull/131
 2025-05-19 20:04:16 +07:00
hathach
67389f37f2 follow up to pr3118, interface also end with IAD. Add more checks 2025-05-16 16:39:53 +07:00
HiFiPhile
531fb69531 Merge pull request #3118 from ZakDanger/vendor_device_fix 
vendor device open fix for descriptor parsing
 2025-05-16 10:25:28 +02:00
hathach
fe4446090e fix dual example for rp2350 conflict printf and cdc_printf 2025-05-13 16:27:26 +07:00
hathach
baf67539fc fix warnings, minor clean up 2025-05-13 12:16:22 +07:00
HiFiPhile
a2117d5d0f simplify vendord_open parsing loop 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-05-09 16:43:19 +02:00
ZakDanger
6b30a61bdb revert 'end fix' 2025-05-09 20:44:11 +10:00
ZakDanger
ae8c00f5d7 revert 'end fix' 2025-05-09 20:42:19 +10:00
ZakDanger
f5944d35b0 vendor device open fix for descriptor parsing 2025-05-09 18:17:07 +10:00
hathach
809af3e74c chore(usbh): change removing_hubs to array instead of bitmask 2025-05-06 15:58:07 +07:00
hathach
bc37ed6e3e usbh: force removed device in the same bus info, before setting address. 
usbh: move code around
hub: queue status endpoint for detach/remove event
 2025-05-06 15:23:23 +07:00
hathach
2abd3c54c4 define hcd_devtree_info_t forr backward compatible 2025-04-29 20:49:47 +07:00
hathach
e7d4b5c9e7 add enum for set addr recovery 2025-04-29 20:16:50 +07:00
hathach
fc43eeddf2 attach debouncing fixed issue with port1 highspeed on imxrt 2025-04-29 10:11:20 +07:00
hathach
42d4f7c81e remove the old attach duplicated logic, debouncing skip should take care of it. 2025-04-28 15:10:03 +07:00
hathach
2c1414b4c1 usbh: add roothub debounncing flag to ignore attach/remove event on the roothub that is currently doing debouncing delay 2025-04-28 14:59:32 +07:00
hathach
b5b7a4be60 hub check status before get 1st device descriptor 2025-04-24 22:21:38 +07:00
hathach
093720f60b fix build 2025-04-24 18:08:00 +07:00
hathach
0f784e8a07 refactor hub driver and move port reset on connection change to usbh. 
hub: add hub_port_get_status_local(), ignore resp in hub_port_get_status(pot != 0)
usbh properly deboucning with hub/rootport accordingly to usb specs, also add 10ms of reset recovery
 2025-04-24 17:59:45 +07:00
hathach
89f8d0cffb add tuh_address_set() API 
minor rename and move code around
 2025-04-24 10:41:00 +07:00
hathach
9a1f690ec4 move usbh ctrl_xfer into usbh_data 2025-04-23 16:50:58 +07:00
hathach
a2da575793 rename and expose tuh_bus_info_get() to application 2025-04-23 16:03:40 +07:00
hathach
741cb3cf02 rename hcd_devtree_info_t to tuh_bus_info_t, hcd_devtree_get_info to hcd_bus_info_get 
streamline bus info to usbh_devies, also replace dev0 (renamed to dev0_bus)
 2025-04-23 12:35:32 +07:00
HiFiPhile
b6170c965f Compile fix. 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-22 21:14:04 +02:00
hathach
8f9ef7dfbe reduce ENUM_DEBOUNCING_DELAY_MS to 200ms 
replace dev0.enumerating by enumerating_daddr for better clean up on unplugging while enumerating
move controller_id & enumerating_daddr into _usbh_data struct
 2025-04-22 22:09:06 +07:00
hathach
940fe43e68 move removing dev0 to process_removing_device() 2025-04-22 17:33:37 +07:00
HiFiPhile
7ba63a6302 Also cleanup unaddressed device. 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-22 16:58:13 +07:00
HiFiPhile
3c4e6a779d Move decouncing delay before USB reset. 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-22 16:58:12 +07:00
HiFiPhile
9f096ac56b host: fix enumerate racing 
- if a previous enumeration failed _ctrl_xfer status could stuck, it needs to be cleared before next attempt.
- after _dev0.enumerating is reset in hcd_event_handler(), if an attach event arrived before _ctrl_xfer clean up in remove event, a racing condition will happen.

Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-22 16:58:10 +07:00
Ha Thach
18d7a993be Merge pull request #3075 from maximevince/dwc2-proper-attach-debouncing 
dwc2/host: attach debouncing fixes
 2025-04-21 22:46:15 +07:00
hathach
5725d33121 improve usbh stability with failed setup send, prevent control stage locked out 2025-04-21 20:39:23 +07:00
Ha Thach
56ed51c8fb Merge pull request #3088 from HiFiPhile/intv 
Fix 1st nak retry one frame shorter.
 2025-04-21 18:31:43 +07:00
HiFiPhile
b3a9b6e37f enable SOF interrupt only if not already enabled 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-19 11:43:28 +02:00
hathach
d51863d1a0 - correctly do_ping if received nyet as transfer complete e.g msc 31 byte command 
- correctly carry out OUT transfer when PING is ack
 2025-04-18 22:46:37 +07:00
HiFiPhile
b3d20442e2 Fix usbh racing later. 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-18 14:57:53 +02:00
HiFiPhile
4c6736c277 Merge branch 'master' into dwc2-proper-attach-debouncing 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-18 14:54:20 +02:00
hathach
8111e53ff0 minor rename 2025-04-18 18:21:42 +07:00
hathach
2fd0301f0e Merge branch 'master' into fork/HiFiPhile/ping_out 
# Conflicts:
#	src/portable/synopsys/dwc2/hcd_dwc2.c
 2025-04-18 17:59:38 +07:00
HiFiPhile
b7a26cc33c Fix 1st nak retry one frame shorter. 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-18 12:46:20 +02:00
Ha Thach
e44f556a95 Merge pull request #3072 from maximevince/dwc2-retry-in-token-immediately 
dwc2/host: immediately retry IN token for bInterval=1
 2025-04-18 17:12:07 +07:00
HiFiPhile
ee1a5dbac0 Merge pull request #3084 from joelpmichael/hcd-template-comments 
note potential issues using ep_desc in hcd_edpt_open()
 2025-04-18 11:25:59 +02:00
HiFiPhile
91163c9217 Merge pull request #3067 from maximevince/dwc2-host-fix-disconnect 
dwc2/host: enable disconnect interrupt + handle it
 2025-04-18 11:24:23 +02:00
hathach
1b888a3311 clean up, remove halted_sof_schedule flags since channel_xfer_in_retry() is only called when channel is halted. 2025-04-18 16:17:35 +07:00
HiFiPhile
b6abc9022a Merge remote-tracking branch 'upstream/master' into hcd-template-comments 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-18 11:13:15 +02:00
HiFiPhile
7134109973 Update hcd_edpt_open() note. 
Signed-off-by: HiFiPhile <admin@hifiphile.com>
 2025-04-18 11:12:14 +02:00