On Wed, 2024-04-10 at 08:17 -0700, Guenter Roeck wrote: > On Mon, Apr 01, 2024 at 01:56:05PM +1100, Frank Crawford wrote: > > Major part of the change for the new method to avoid chipset > > issues. > > > > Perform an intial test if the chip ID can be read without entering > > configuration mode, if so, do not enter configuration mode and > > continue > > as is. > > > > If chip ID cannot be read, enter configuration mode, and continue > > with > > previous code. > > > > Also update exit code to take account of if we entered > > configuration > > mode or not. > > > > You describe the changes, but you don't describe the problem you are > trying to solve. Even if configuration mode was already entered, that > is not a reason to keep it active. We don't _know_ what is going on > outside the driver and can not make assumptions. For changes like > this > you really need to explain the problem you are trying to solve, and > the > reasoning behind the changes. Just assuming that a set of chips would > have its SIO mode enabled by the BIOS is just wrong. We don't know > what if anything the BIOS is doing. In this regard, it is no different to the current code, which runs SIO mode enable, but does not ever run the SIO mode disable code. In fact this code is actually safer than the previous code in that it acts similar to the chip not being enabled or disabled, which would happen if no driver existed. > > Guenter > Regards Frank
