usb: cdnsp: Fix deadlock issue during using NCM gadget

The interrupt service routine registered for the gadget is a primary
handler which mask the interrupt source and a threaded handler which
handles the source of the interrupt. Since the threaded handler is
voluntary threaded, the IRQ-core does not disable bottom halves before
invoke the handler like it does for the forced-threaded handler.

Due to changes in networking it became visible that a network gadget's
completions handler may schedule a softirq which remains unprocessed.
The gadget's completion handler is usually invoked either in hard-IRQ or
soft-IRQ context. In this context it is enough to just raise the softirq
because the softirq itself will be handled once that context is left.
In the case of the voluntary threaded handler, there is nothing that
will process pending softirqs. Which means it remain queued until
another random interrupt (on this CPU) fires and handles it on its exit
path or another thread locks and unlocks a lock with the bh suffix.
Worst case is that the CPU goes idle and the NOHZ complains about
unhandled softirqs.

Disable bottom halves before acquiring the lock (and disabling
interrupts) and enable them after dropping the lock. This ensures that
any pending softirqs will handled right away.

cc: stable@vger.kernel.org
Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Acked-by: Peter Chen <peter.chen@kernel.org>
Link: https://lore.kernel.org/r/20231108093125.224963-1-pawell@cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb:cdnsp: remove TRB_FLUSH_ENDPOINT command

Patch removes TRB_FLUSH_ENDPOINT command from driver.
This command is not supported by controller and
USBSSP returns TRB Error completion code for it.

Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Acked-by: Peter Chen <peter.chen@kernel.org>
Link: https://lore.kernel.org/r/20231026073737.165450-1-pawell@cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: : add scatter gather support for ISOC endpoint

Patch implements scatter gather support for isochronous endpoint.
This fix is forced by 'commit e81e7f9a0eb9
("usb: gadget: uvc: add scatter gather support")'.
After this fix CDNSP driver stop working with UVC class.

Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Link: https://lore.kernel.org/r/20221222090934.145140-1-pawell@cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: fix lack of ZLP for ep0

Patch implements the handling of ZLP for control transfer.
To send the ZLP driver must prepare the extra TRB in TD with
length set to zero and TRB type to TRB_NORMAL.
The first TRB must have set TRB_CHAIN flag, TD_SIZE = 1
and TRB type to TRB_DATA.

Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
cc: <stable@vger.kernel.org>
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Link: https://lore.kernel.org/r/20221122085138.332434-1-pawell@cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: fix issue with ZLP - added TD_SIZE = 1

Patch modifies the TD_SIZE in TRB before ZLP TRB.
The TD_SIZE in TRB before ZLP TRB must be set to 1 to force
processing ZLP TRB by controller.

cc: <stable@vger.kernel.org>
Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Link: https://lore.kernel.org/r/20221115092218.421267-1-pawell@cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: Fix issue with Clear Feature Halt Endpoint

During handling Clear Halt Endpoint Feature request, driver invokes
Reset Endpoint command. Because this command has some issue with
transition endpoint from Running to Idle state the driver must
stop the endpoint by using Stop Endpoint command.

cc: <stable@vger.kernel.org>
Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Link: https://lore.kernel.org/r/20221110063005.370656-1-pawell@cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: Fixed setting last_trb incorrectly

When ZLP occurs in bulk transmission, currently cdnsp will set last_trb
for the last two TRBs, it will trigger an error "ERROR Transfer event TRB
DMA ptr not part of current TD ...".

Fixes: e913aada0683 ("usb: cdnsp: Fixed issue with ZLP")
Cc: stable <stable@kernel.org>
Acked-by: Pawel Laszczak <pawell@cadence.com>
Signed-off-by: Jing Leng <jleng@ambarella.com>
Link: https://lore.kernel.org/r/20220609021134.1606-1-3090101217@zju.edu.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: Fix incorrect status for control request

Patch fixes incorrect status for control request.
Without this fix all usb_request objects were returned to upper drivers
with usb_reqest->status field set to -EINPROGRESS.

Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
cc: <stable@vger.kernel.org>
Reported-by: Ken (Jian) He <jianhe@ambarella.com>
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Link: https://lore.kernel.org/r/20211207091838.39572-1-pawell@gli-login.cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: Fix incorrect calling of cdnsp_died function

Patch restrict calling of cdnsp_died function during removing modules
or software disconnect.
This function was called because after transition controller to HALT
state the driver starts handling the deferred interrupt.
In this case such interrupt can be simple ignored.

Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
cc: <stable@vger.kernel.org>
Reviewed-by: Peter Chen <peter.chen@kernel.org>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Link: https://lore.kernel.org/r/20211210112945.660-1-pawell@gli-login.cadence.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: cdnsp: Fixed issue with ZLP

The condition "if (need_zero_pkt && zero_len_trb)" was always false
and it caused that TRB for ZLP was not prepared.

Fix causes that after preparing last TRB in TD, the driver prepares
additional TD with ZLP when a ZLP is required.

Cc: <stable@vger.kernel.org>
Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Link: https://lore.kernel.org/r/20210623072728.41275-1-pawell@gli-login.cadence.com
Signed-off-by: Peter Chen <peter.chen@kernel.org>

usb: cdnsp: Fix deadlock issue in cdnsp_thread_irq_handler

Patch fixes the following critical issue caused by deadlock which has been
detected during testing NCM class:

smp: csd: Detected non-responsive CSD lock (#1) on CPU#0
smp: csd: CSD lock (#1) unresponsive.
....
RIP: 0010:native_queued_spin_lock_slowpath+0x61/0x1d0
RSP: 0018:ffffbc494011cde0 EFLAGS: 00000002
RAX: 0000000000000101 RBX: ffff9ee8116b4a68 RCX: 0000000000000000
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff9ee8116b4658
RBP: ffffbc494011cde0 R08: 0000000000000001 R09: 0000000000000000
R10: ffff9ee8116b4670 R11: 0000000000000000 R12: ffff9ee8116b4658
R13: ffff9ee8116b4670 R14: 0000000000000246 R15: ffff9ee8116b4658
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f7bcc41a830 CR3: 000000007a612003 CR4: 00000000001706e0
Call Trace:
<IRQ>
do_raw_spin_lock+0xc0/0xd0
_raw_spin_lock_irqsave+0x95/0xa0
cdnsp_gadget_ep_queue.cold+0x88/0x107 [cdnsp_udc_pci]
usb_ep_queue+0x35/0x110
eth_start_xmit+0x220/0x3d0 [u_ether]
ncm_tx_timeout+0x34/0x40 [usb_f_ncm]
? ncm_free_inst+0x50/0x50 [usb_f_ncm]
__hrtimer_run_queues+0xac/0x440
hrtimer_run_softirq+0x8c/0xb0
__do_softirq+0xcf/0x428
asm_call_irq_on_stack+0x12/0x20
</IRQ>
do_softirq_own_stack+0x61/0x70
irq_exit_rcu+0xc1/0xd0
sysvec_apic_timer_interrupt+0x52/0xb0
asm_sysvec_apic_timer_interrupt+0x12/0x20
RIP: 0010:do_raw_spin_trylock+0x18/0x40
RSP: 0018:ffffbc494138bda8 EFLAGS: 00000246
RAX: 0000000000000000 RBX: ffff9ee8116b4658 RCX: 0000000000000000
RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff9ee8116b4658
RBP: ffffbc494138bda8 R08: 0000000000000001 R09: 0000000000000000
R10: ffff9ee8116b4670 R11: 0000000000000000 R12: ffff9ee8116b4658
R13: ffff9ee8116b4670 R14: ffff9ee7b5c73d80 R15: ffff9ee8116b4000
_raw_spin_lock+0x3d/0x70
? cdnsp_thread_irq_handler.cold+0x32/0x112c [cdnsp_udc_pci]
cdnsp_thread_irq_handler.cold+0x32/0x112c [cdnsp_udc_pci]
? cdnsp_remove_request+0x1f0/0x1f0 [cdnsp_udc_pci]
? cdnsp_thread_irq_handler+0x5/0xa0 [cdnsp_udc_pci]
? irq_thread+0xa0/0x1c0
irq_thread_fn+0x28/0x60
irq_thread+0x105/0x1c0
? __kthread_parkme+0x42/0x90
? irq_forced_thread_fn+0x90/0x90
? wake_threads_waitq+0x30/0x30
? irq_thread_check_affinity+0xe0/0xe0
kthread+0x12a/0x160
? kthread_park+0x90/0x90
ret_from_fork+0x22/0x30

The root cause of issue is spin_lock/spin_unlock instruction instead
spin_lock_irqsave/spin_lock_irqrestore in cdnsp_thread_irq_handler
function.

Cc: stable@vger.kernel.org
Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
Signed-off-by: Pawel Laszczak <pawell@cadence.com>

Link: https://lore.kernel.org/r/20210526060527.7197-1-pawell@gli-login.cadence.com
Signed-off-by: Peter Chen <peter.chen@kernel.org>

usb: cdnsp: Fixes incorrect value in ISOC TRB

Fixes issue with priority of operator. Operator "|" priority is
higher then "? :".
To improve the readability the operator "? :" has been replaced with
"if ()" statement.

Fixes: 3d82904559f4 ("usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver")
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Signed-off-by: Peter Chen <peter.chen@kernel.org>

usb: cdnsp: Removes some useless trace events

Patch removes some useless trace events that can
be replaced by ftrace.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Signed-off-by: Peter Chen <peter.chen@kernel.org>

usb: cdnsp: Fixes for sparse warnings

Patch fixes all sparse warnings in cdsnp driver.

It fixes the following warnings:
cdnsp-ring.c:1441: warning: incorrect type in assignment
cdnsp-ring.c:1444: warning: restricted __le32 degrades to integer
cdnsp-ring.c:2200: warning: dubious: x | !y
cdnsp-gadget.c:501: warning: incorrect type in assignment
cdnsp-gadget.c:504: warning: restricted __le32 degrades to integer
cdnsp-gadget.c:507: warning: restricted __le32 degrades to integer
cdnsp-gadget.c:508: warning: restricted __le32 degrades to integer
cdnsp-gadget.c:509: warning: invalid assignment: |=
cdnsp-gadget.c:510: warning: cast from restricted __le32
cdnsp-gadget.c:558: warning: incorrect type in assignment
cdnsp-gadget.c:561: warning: restricted __le32 degrades to integer
cdnsp-gadget.c:570: warning: restricted __le32 degrades to integer
cdnsp-gadget.c:1571: warning: incorrect type in argument 1
cdnsp-gadget.c:1602: warning: restricted __le32 degrades to integer
cdnsp-gadget.c:1760: warning: incorrect type in assignment
cdnsp-gadget.c:1762: warning: incorrect type in assignment
cdnsp-gadget.c:1763: warning: incorrect type in assignment
cdnsp-gadget.c:1764: warning: incorrect type in assignment
cdnsp-gadget.c:1765: warning: incorrect type in assignment
cdnsp-gadget.c:1766: warning: incorrect type in assignment
cdnsp-gadget.c:1767: warning: incorrect type in assignment
cdnsp-gadget.c:458: warning: cast truncates bits from constant value
(ffffffff07ffffff becomes 7ffffff)
cdnsp-gadget.c:666: warning: cast truncates bits from constant value
(ffffffff07ffffff becomes 7ffffff)
cdnsp-mem.c:762: warning: incorrect type in assignment
cdnsp-mem.c:763: warning: incorrect type in assignment
cdnsp-mem.c:928: warning: cast from restricted __le16
cdnsp-mem.c:1187: warning: incorrect type in assignment
cdnsp-mem.c:1191: warning: incorrect type in assignment
cdnsp-ep0.c:142: warning: incorrect type in assignment
cdnsp-ep0.c:144: warning: restricted __le32 degrades to integer
cdnsp-ep0.c:147: warning: restricted __le32 degrades to integer
cdnsp-ep0.c:148: warning: restricted __le32 degrades to integer
cdnsp-ep0.c:179: warning: incorrect type in argument 1
cdnsp-ep0.c:311: warning: incorrect type in argument 1
cdnsp-ep0.c:469: warning: incorrect type in assignment
cdnsp-trace.h:611:1: warning: cast from restricted __le32

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>

usb: cdnsp: Add tracepoints for CDNSP driver

Patch adds the series of tracepoints that can be used for
debugging issues detected in driver.

Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Reviewed-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>

usb: cdnsp: cdns3 Add main part of Cadence USBSSP DRD Driver

This patch introduces the main part of Cadence USBSSP DRD driver
to Linux kernel.
To reduce the patch size a little bit, the header file gadget.h was
intentionally added as separate patch.

The Cadence USBSSP DRD Controller is a highly configurable IP Core which
can be instantiated as Dual-Role Device (DRD), Peripheral Only and
Host Only (XHCI)configurations.

The current driver has been validated with FPGA platform. We have
support for PCIe bus, which is used on FPGA prototyping.

The host side of USBSS DRD controller is compliant with XHCI.
The architecture for device side is almost the same as for host side,
and most of the XHCI specification can be used to understand how
this controller operates.

Signed-off-by: Pawel Laszczak <pawell@cadence.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>