Quoting r. Richard B. Johnson <jmodem@xxxxxxxxxxxxxxxxxxxxx>: > Subject: Re: Ordering between PCI config space writes and MMIO reads? > > > ----- Original Message ----- > From: "Michael S. Tsirkin" <mst@xxxxxxxxxxxxxx> > To: "Roland Dreier" <rdreier@xxxxxxxxx> > Cc: <linux-kernel@xxxxxxxxxxxxxxx>; <linux-ia64@xxxxxxxxxxxxxxx>; > <jeff@xxxxxxxxxx>; <matthew@xxxxxx>; <openib-general@xxxxxxxxxx>; > <linux-pci@xxxxxxxxxxxxxxxxxxxxxxxx>; "David Miller" <davem@xxxxxxxxxxxxx> > Sent: Tuesday, October 31, 2006 2:53 PM > Subject: Re: Ordering between PCI config space writes and MMIO reads? > > > > Quoting r. Roland Dreier <rdreier@xxxxxxxxx>: > >> Subject: Re: Ordering between PCI config space writes and MMIO reads? > >> > >> The discussion fizzled out without really reaching a definitive > >> answer, so I'm going to apply the original patch (below), since I > >> pretty much convinced myself that only the driver doing the config > >> access has enough information to fix this reliably. > >> > >> - R. > >> > >> Author: John Partridge <johnip@xxxxxxx> > >> Date: Tue Oct 31 11:00:04 2006 -0800 > >> > >> IB/mthca: Make sure all PCI config writes reach device before doing > >> MMIO > >> > >> During initialization, mthca writes some PCI config space registers > >> and then does an MMIO read from one of the BARs it just enabled. > >> This > >> MMIO read sometimes failed and caused a crash on SGI Altix machines, > >> because the PCI-X host bridge (legitimately, according to the PCI > >> spec) allowed the MMIO read to start before the config write > >> completed. > >> > >> To fix this, add a config read after all config writes to make sure > >> they are all done before starting the MMIO read. > >> > >> Signed-off-by: John Partridge <johnip@xxxxxxx> > >> Signed-off-by: Roland Dreier <rolandd@xxxxxxxxx> > >> > >> diff --git a/drivers/infiniband/hw/mthca/mthca_reset.c > >> b/drivers/infiniband/hw/mthca/mthca_reset.c > >> index 91934f2..578dc7c 100644 > >> --- a/drivers/infiniband/hw/mthca/mthca_reset.c > >> +++ b/drivers/infiniband/hw/mthca/mthca_reset.c > >> @@ -281,6 +281,20 @@ good: > >> goto out; > >> } > >> > >> + /* > >> + * Perform a "flush" of the PCI config writes here by reading > >> + * the PCI_COMMAND register. This is needed to make sure that > >> + * we don't try to touch other PCI BARs before the config > >> + * writes are done -- otherwise an MMIO cycle could start > >> + * before the config writes are done and reach the HCA before > >> + * the BAR is actually enabled. > >> + */ > >> + if (pci_read_config_dword(mdev->pdev, PCI_COMMAND, hca_header)) { > >> + err = -ENODEV; > >> + mthca_err(mdev, "Couldn't access HCA memory after restoring, " > >> + "aborting.\n"); > >> + } > >> + > >> out: > >> if (bridge) > >> pci_dev_put(bridge); > > > > Here's what I don't understand: according to PCI rules, pci config read > > can bypass pci config write (both are non-posted). > > So why does doing it help flush the writes as the comment claims? > > > > Isn't this more the case of > > /* pci_config_write seems to complete asynchronously on Altix systems. > > * This is probably broken but its not clear what's the best > > * thing to do is - for now, do pci_read_config_dword which seems to flush > > * everything out. */ > > > > If you write to the PCI bus and then you read the result, the read __might__ > be the read that flushes any posted writes rather than the read of device > registers that would occur after the BARs were configured (hardware may be > slower than the CPU). So, it's best to do the required configuration cycles > first, then after all is done, read something before you actually need to use > data from subsequent read/write cycles. But why should it help? Accordig to the spec, read does not flush configuration writes (unlike regular writes). -- MST - To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
![]() |