Re: [PATCH v4 1/5] vpci: fix updating the command register

From: "Roger Pau Monné" <roger.pau@citrix.com>
To: Jan Beulich <JBeulich@suse.com>
Cc: Stefano Stabellini <sstabellini@kernel.org>,
	Wei Liu <wei.liu2@citrix.com>,
	Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>,
	George Dunlap <George.Dunlap@eu.citrix.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	Ian Jackson <Ian.Jackson@eu.citrix.com>, Tim Deegan <tim@xen.org>,
	Julien Grall <julien.grall@arm.com>,
	xen-devel <xen-devel@lists.xenproject.org>
Subject: Re: [PATCH v4 1/5] vpci: fix updating the command register
Date: Fri, 16 Nov 2018 15:32:45 +0100	[thread overview]
Message-ID: <20181116143245.vpk3vgpbcqujvofb@mac> (raw)
In-Reply-To: <5BEEB15D02000078001FCB9A@prv1-mh.provo.novell.com>

On Fri, Nov 16, 2018 at 05:00:29AM -0700, Jan Beulich wrote:
> >>> On 14.11.18 at 12:57, <roger.pau@citrix.com> wrote:
> > @@ -413,7 +412,7 @@ static void rom_write(const struct pci_dev *pdev, unsigned int reg,
> >          header->rom_enabled = new_enabled;
> >          pci_conf_write32(pdev->seg, pdev->bus, slot, func, reg, val);
> >      }
> > -    else if ( modify_bars(pdev, new_enabled, true) )
> > +    else if ( modify_bars(pdev, new_enabled ? PCI_COMMAND_MEMORY : 0, true) )
> 
> Do you really mean to clear all other defined bits of the command
> register here?

This is a ROM BAR write, not a command register write. rom_write
passes PCI_COMMAND_MEMORY merely to signal this is a mapping
operation, but the value would never be written to the command
register, there's an ASSERT(!rom_only) just before the deferred write
of the command register in modify_decoding.

Thanks, Roger.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel