Re: [PATCH] wifi: brcm80211: move the USB bootloader definitions to a shared header
From: Ragul Balaji
Date: Tue Nov 17 2026 - 17:12:06 EST
On Tue, Nov 17, 2026 at 6:52 AM Arend van Spriel
<arend.vanspriel@xxxxxxxxxxxx> wrote:
>
> On Thu, 12 Nov 2026 13:08:15 +0800, Ragul Balaji wrote:
> > The Broadcom USB bootloader downloads firmware the same way for other
> > USB chips than the FullMAC ones of brcmfmac. Move its commands, states,
> > firmware header and response layouts from brcmfmac's usb.c to a header
> > in the shared include directory, so that other drivers can use them.
> >
> > Signed-off-by: Ragul Balaji <ragulbaby@xxxxxxxxx>
> > Signed-off-by: Ho Jie Feng <hjf3108@xxxxxxxxx>
> > ---
> > .../net/wireless/broadcom/brcm80211/brcmfmac/usb.c | 68 +---------------
> > .../broadcom/brcm80211/include/brcm_usb_rdl.h | 78 ++++++++++++++++++++++
> > 2 files changed, 79 insertions(+), 67 deletions(-)
>
> [...]
>
> > +#ifndef _BRCM_USB_RDL_H_
> > +#define _BRCM_USB_RDL_H_
> > +
> > +#include <linux/types.h>
> > +
> > +#define BRCMF_POSTBOOT_ID 0xA123 /* ID to detect if dongle
> > + has boot up */
> > +
> > +#define TRX_MAGIC 0x30524448 /* "HDR0" */
>
> [...]
>
> The move itself looks fine. Like the series "Move brcmsmac code into
> brcmutil in preparation for brcmbmac" (posted a few minutes later), it is
> part of the reorganization that has to go in before the brcmbmac driver
> is sent again, and nothing else in the tree uses this header yet. Keeping
> the brcmfmac and the brcmsmac parts as separate patches is fine, but
> please send them as one series with a common cover letter, so the shared
> pieces are reviewed together. Since the series then touches more than
> brcmsmac, rename it to something like "Reorganize brcm80211 code into
Thanks for the review and suggestions. I'll combine the brcmfmac and
brcmsmac changes into a single patch series with a common cover
letter, and update the series title as suggested.
> shared directories in preparation for brcmbmac". Can you clarify the
> additional sign-off from Ho Jie Feng? Did he co-author the patch or did
> you use his (earlier) work?
Regarding Ho Jie Feng's sign-off, this is my first time working on
this kind of contribution, and Jie Feng has been guiding me through
the process. We've worked on this together, including the
implementation, hardware setup, and review. His contributions have
been significant throughout the process, which is why I've included
his sign-off.
>
> Regards,
> Arend