[PATCH] Module argument to control whether intel-spi-pci attempts to turn the SPI flash chip writeable

[Daniel Gutson]

On Thu, Jul 30, 2020 at 11:18 AM Daniel Gutson <daniel at eclypsium.com> wrote:
>
> On Thu, Jul 30, 2020 at 11:09 AM Arnd Bergmann <arnd at arndb.de> wrote:
> >
> > On Thu, Jul 30, 2020 at 2:21 PM Daniel Gutson <daniel at eclypsium.com> wrote:
> > > El jue., 30 jul. 2020 2:31 a. m., Greg Kroah-Hartman <gregkh at linuxfoundation.org> escribió:
> > >>
> > >> Again, module parameters are working on a per-chunk-of-code basis, while
> > >> you want to work on a per-device basis,
> > >
> > >
> > > I think there is a misunderstanding.  What I want is to control (turn on or off) is a very specific code snippet that provides the "functionality" of trying to turn the chip writable. The rest of the device driver is fine.
> > > I assume that the one that doesn't understand is me.
> > >
> >
> > I looked at the source code again and found that the existing module
> > parameter applies to both the platform and pci device front-ends, both
> > of which go through
> >
> >         /* Prevent writes if not explicitly enabled */
> >         if (!ispi->writeable || !writeable)
> >                 ispi->nor.mtd.flags &= ~MTD_WRITEABLE;
> >
>
> I think you missed
> https://elixir.bootlin.com/linux/latest/source/drivers/mtd/spi-nor/controllers/intel-spi-pci.c#L44
>
>     /* Try to make the chip read/write */
>     pci_read_config_dword(pdev, BCR, &bcr);
>     if (!(bcr & BCR_WPD)) {
>         bcr |= BCR_WPD;
>         pci_write_config_dword(pdev, BCR, bcr);
>         pci_read_config_dword(pdev, BCR, &bcr);
>     }
>
> in the probe function, and is executed always and unconditionally.

To clarify, this is executed before intel-spi code.

>
> /* Try to make the chip read/write */
> pci_read_config_dword(pdev, BCR, &bcr);
> if (!(bcr & BCR_WPD)) {
> bcr |= BCR_WPD;
> pci_write_config_dword(pdev, BCR, bcr);
> pci_read_config_dword(pdev, BCR, &bcr);
> }
>
> > Setting the PCI device writable in hardware makes it possible to
> > actually write to it *only* if the module parameter is also set to '1'.
> > One might disagree with that design, but I don't think your patch
> > would make it any better, it just means one would have to set
> > two module parameters instead of one.
> >
> >      Arnd
>
>
>
> --
> Daniel Gutson
> Argentina Site Director
> Enginieering Director
> Eclypsium
>
> Below The Surface: Get the latest threat research and insights on
> firmware and supply chain threats from the research team at Eclypsium.
> https://eclypsium.com/research/#threatreport



-- 
Daniel Gutson
Argentina Site Director
Enginieering Director
Eclypsium

Below The Surface: Get the latest threat research and insights on
firmware and supply chain threats from the research team at Eclypsium.
https://eclypsium.com/research/#threatreport