On Mon, 8 Nov 2021 18:20:43 -0800 Roman Bacik roman.bacik@broadcom.com wrote:
On Mon, Nov 8, 2021 at 5:12 PM Marek Behún kabel@kernel.org wrote:
On Mon, 8 Nov 2021 16:48:33 -0800 Roman Bacik roman.bacik@broadcom.com wrote:
To be honest changing status codes coming from FW does not seem right. But we will try to make the requested changes.
I looked at kernel's implementation of this driver and these hwrm functions and they don't return STATUS_*.
Marek
Marek,
This is quite a different driver and it was written for uboot.
Hello Roman
The drivers clearly have a common ancestor, there are far too many similarities. It clearly wasn't written from scratch for U-Boot.
If the main objection is that Linux driver is different then maybe we should use v10 as is. Currently hwrm methods return HW status and bnxt methods return uboot error codes consistently.
I will leave this to U-Boot's network subsystem maintainers. As I said, beggars cannot be choosers in U-Boot. As long as the driver does not introduce vendor specific stuff to the user API (in U-Boot command line), then I guess I'll have to be satisfied.
Marek