Thread (18 messages) 18 messages, 5 authors, 2025-02-20

Re: [PATCH net-next 2/2] net: mctp: Add MCTP USB transport driver

From: Jeremy Kerr <jk@codeconstruct.com.au>
Date: 2025-02-07 09:45:35
Also in: linux-usb

Hi Greg,
quoted
quoted
quoted
+               dev_err(&mctp_usb->usbdev->dev, "%s: urb status: %d\n",
+                       __func__, status);
This could flood the logs, are you sure you need it at dev_err()
level?

And __func__ is redundant, it's present in dev_*() calls already.
am I missing something then?

   [  146.130170] usb 2-1: short packet (hdr) 6

emitted from:

    dev_dbg(&mctp_usb->usbdev->dev,
            "short packet (hdr) %d\n",
            hdr->len);

Seems like we get the driver name, but not the function.

I'm happy to remove the __func__ output either way, but I will also
make the logs a little more descriptive for context, if we don't have
func data.
Please read Documentation/admin-guide/dynamic-debug-howto.rst, it shows
how to get the function information from the dev_dbg() lines at runtime.

In short:
        $ alias ddcmd='echo $* > /proc/dynamic_debug/control'
        # add function to all enabled messages
        $ ddcmd '+f'
Your original comment was on the dev_err() call though (sorry, I've
complicated the discussion by using a dev_dbg() example).

Looks like only dev_dbg (and not _err/_warn/etc) has provision for
__func__, is that right?

I've since removed the __func__ references anyway, and replaced with
better context on the messages, but keen to make sure I have the correct
understanding in general.

Cheers,


Jeremy
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help