Re: [PATCH 3/4] usb: cdns2: Add tracepoints for CDNS2 driver
From: Greg KH
Date: Thu Feb 16 2023 - 06:23:00 EST
On Thu, Feb 16, 2023 at 05:54:10AM -0500, Pawel Laszczak wrote:
> Patch adds the series of tracepoints that can be used for
> debugging issues detected in driver.
>
> Signed-off-by: Pawel Laszczak <pawell@xxxxxxxxxxx>
> ---
> drivers/usb/gadget/udc/cdns2/Makefile | 2 +
> drivers/usb/gadget/udc/cdns2/cdns2-debug.h | 200 +++++++
> drivers/usb/gadget/udc/cdns2/cdns2-ep0.c | 26 +-
> drivers/usb/gadget/udc/cdns2/cdns2-gadget.c | 52 +-
> drivers/usb/gadget/udc/cdns2/cdns2-trace.c | 11 +
> drivers/usb/gadget/udc/cdns2/cdns2-trace.h | 609 ++++++++++++++++++++
> 6 files changed, 897 insertions(+), 3 deletions(-)
> create mode 100644 drivers/usb/gadget/udc/cdns2/cdns2-debug.h
> create mode 100644 drivers/usb/gadget/udc/cdns2/cdns2-trace.c
> create mode 100644 drivers/usb/gadget/udc/cdns2/cdns2-trace.h
>
> diff --git a/drivers/usb/gadget/udc/cdns2/Makefile b/drivers/usb/gadget/udc/cdns2/Makefile
> index 7c746e6d53c2..a1ffbbe2e768 100644
> --- a/drivers/usb/gadget/udc/cdns2/Makefile
> +++ b/drivers/usb/gadget/udc/cdns2/Makefile
> @@ -1,5 +1,7 @@
> # SPDX-License-Identifier: GPL-2.0
> # define_trace.h needs to know how to find our header
> +CFLAGS_cdns2-trace.o := -I$(src)
>
> obj-$(CONFIG_USB_CDNS2_UDC) += cdns2-udc-pci.o
> cdns2-udc-pci-$(CONFIG_USB_CDNS2_UDC) += cdns2-pci.o cdns2-gadget.o cdns2-ep0.o
> +cdns2-udc-pci-$(CONFIG_TRACING) += cdns2-trace.o
> diff --git a/drivers/usb/gadget/udc/cdns2/cdns2-debug.h b/drivers/usb/gadget/udc/cdns2/cdns2-debug.h
> new file mode 100644
> index 000000000000..3a80f3fbdcd3
> --- /dev/null
> +++ b/drivers/usb/gadget/udc/cdns2/cdns2-debug.h
> @@ -0,0 +1,200 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +/*
> + * Cadence USBHS-DEV Driver.
> + * Debug header file.
> + *
> + * Copyright (C) 2023 Cadence.
> + *
> + * Author: Pawel Laszczak <pawell@xxxxxxxxxxx>
> + */
> +
> +#ifndef __LINUX_CDNS2_DEBUG
> +#define __LINUX_CDNS2_DEBUG
> +
> +static inline const char *cdns2_decode_usb_irq(char *str, size_t size,
> + u8 usb_irq, u8 ext_irq)
<snip>
Why are all of these big functions declared as inline? Do they really
work as inline functions if you only reference them in a trace callback?
thanks,
greg k-h