From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([140.186.70.92]:53057) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1S01Ob-00006C-Qh for qemu-devel@nongnu.org; Tue, 21 Feb 2012 20:53:30 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1S01Oa-0002Wg-HR for qemu-devel@nongnu.org; Tue, 21 Feb 2012 20:53:29 -0500 Received: from gate.crashing.org ([63.228.1.57]:33766) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1S01Oa-0002WM-3L for qemu-devel@nongnu.org; Tue, 21 Feb 2012 20:53:28 -0500 Message-ID: <1329875595.3980.84.camel@pasglop> From: Benjamin Herrenschmidt Date: Wed, 22 Feb 2012 12:53:15 +1100 In-Reply-To: <4F43B084.6060402@suse.de> References: <20120221135758.GA8102@redhat.com> <4F43B084.6060402@suse.de> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH] pci: set memory type for memory behind the bridge List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Alexander Graf Cc: blauwirbel@gmail.com, qemu-devel@nongnu.org, Gerd Hoffmann , afaerber@suse.de, "Michael S. Tsirkin" On Tue, 2012-02-21 at 15:56 +0100, Alexander Graf wrote: > On 02/21/2012 02:57 PM, Michael S. Tsirkin wrote: > > As we make upper bits in IO and prefetcheable memory > > registers writeable, we should declare support > > for 64 bit prefetcheable memory and 32 bit io > > in the bridge. > > > > This changes the default for apb, dec, but I'm guessing > > they got the defaults wrong by accident. > > Alternatively, we could let bridges declare lack of > > 64 bit support and make the upper bits read-only zero. > > > > With this applied, we can drop these bits > > from express code. > > Ben, you know these devices better than us. Any comments? :) I don't know that code very well, but Michael's explanation at least makes sense to me :-) Cheers, Ben. > Alex > > > Reported-by: Gerd Hoffmann > > Signed-off-by: Michael S. Tsirkin > > > > Could someone familiar with apb,dec ack this please? > > > > --- > > hw/pci.c | 15 +++++++++++++-- > > 1 files changed, 13 insertions(+), 2 deletions(-) > > > > diff --git a/hw/pci.c b/hw/pci.c > > index 31d6a5f..3ca5f4c 100644 > > --- a/hw/pci.c > > +++ b/hw/pci.c > > @@ -611,7 +611,7 @@ static void pci_init_w1cmask(PCIDevice *dev) > > PCI_STATUS_SIG_SYSTEM_ERROR | PCI_STATUS_DETECTED_PARITY); > > } > > > > -static void pci_init_wmask_bridge(PCIDevice *d) > > +static void pci_init_mask_bridge(PCIDevice *d) > > { > > /* PCI_PRIMARY_BUS, PCI_SECONDARY_BUS, PCI_SUBORDINATE_BUS and > > PCI_SEC_LETENCY_TIMER */ > > @@ -632,6 +632,14 @@ static void pci_init_wmask_bridge(PCIDevice *d) > > /* PCI_PREF_BASE_UPPER32 and PCI_PREF_LIMIT_UPPER32 */ > > memset(d->wmask + PCI_PREF_BASE_UPPER32, 0xff, 8); > > > > + /* Supported memory and i/o types */ > > + d->config[PCI_IO_BASE] |= PCI_IO_RANGE_TYPE_32; > > + d->config[PCI_IO_LIMIT] |= PCI_IO_RANGE_TYPE_32; > > + pci_word_test_and_set_mask(d->config + PCI_PREF_MEMORY_BASE, > > + PCI_PREF_RANGE_TYPE_64); > > + pci_word_test_and_set_mask(d->config + PCI_PREF_MEMORY_LIMIT, > > + PCI_PREF_RANGE_TYPE_64); > > + > > /* TODO: add this define to pci_regs.h in linux and then in qemu. */ > > #define PCI_BRIDGE_CTL_VGA_16BIT 0x10 /* VGA 16-bit decode */ > > #define PCI_BRIDGE_CTL_DISCARD 0x100 /* Primary discard timer */ > > @@ -654,6 +662,9 @@ static void pci_init_wmask_bridge(PCIDevice *d) > > * completeness. */ > > pci_set_word(d->w1cmask + PCI_BRIDGE_CONTROL, > > PCI_BRIDGE_CTL_DISCARD_STATUS); > > + d->cmask[PCI_IO_BASE] |= PCI_IO_RANGE_TYPE_MASK; > > + pci_word_test_and_set_mask(d->cmask + PCI_PREF_MEMORY_BASE, > > + PCI_PREF_RANGE_TYPE_MASK); > > } > > > > static int pci_init_multifunction(PCIBus *bus, PCIDevice *dev) > > @@ -775,7 +786,7 @@ static PCIDevice *do_pci_register_device(PCIDevice *pci_dev, PCIBus *bus, > > pci_init_wmask(pci_dev); > > pci_init_w1cmask(pci_dev); > > if (pc->is_bridge) { > > - pci_init_wmask_bridge(pci_dev); > > + pci_init_mask_bridge(pci_dev); > > } > > if (pci_init_multifunction(bus, pci_dev)) { > > pci_config_free(pci_dev);