diff options
| author | jhb <jhb@FreeBSD.org> | 2003-12-31 16:56:32 +0000 |
|---|---|---|
| committer | jhb <jhb@FreeBSD.org> | 2003-12-31 16:56:32 +0000 |
| commit | 390a3f79d84ed94330c5abc29572f173803f05b7 (patch) | |
| tree | 6e9f55b7420eaf55d4749cb1d9cdcbfae4f0afb5 /sys/i386/pci | |
| parent | ec25118d4b47b64a5b8c890a5e3cbd3f82c93b6c (diff) | |
| download | FreeBSD-src-390a3f79d84ed94330c5abc29572f173803f05b7.zip FreeBSD-src-390a3f79d84ed94330c5abc29572f173803f05b7.tar.gz | |
Replace an outb() during the test for configuration mechanism #1 with a
DELAY(1) instead. After wading through old commit logs, I found that the
outb() was added not as part of the test but as an intentional delay. In
fact, according to Shanley's PCI book, the configuration 1 data and address
ports should only be accessed using aligned 32-bit accesses (i.e. inl()
and outl()). Thus, using outb() to just the last byte of the port violates
the PCI spec it would seem. On at least one box doing so broke the probe
for PCI, whereas changing it to a DELAY(1) fixed the probe.
Reported by: Sean Welch <welchsm@earthlink.net>
MFC after: 1 week
Diffstat (limited to 'sys/i386/pci')
| -rw-r--r-- | sys/i386/pci/pci_cfgreg.c | 2 | ||||
| -rw-r--r-- | sys/i386/pci/pci_pir.c | 2 |
2 files changed, 2 insertions, 2 deletions
diff --git a/sys/i386/pci/pci_cfgreg.c b/sys/i386/pci/pci_cfgreg.c index 91746fe..0e3b8c3 100644 --- a/sys/i386/pci/pci_cfgreg.c +++ b/sys/i386/pci/pci_cfgreg.c @@ -732,7 +732,7 @@ pcireg_cfgopen(void) devmax = 32; outl(CONF1_ADDR_PORT, CONF1_ENABLE_CHK); - outb(CONF1_ADDR_PORT + 3, 0); + DELAY(1); mode1res = inl(CONF1_ADDR_PORT); outl(CONF1_ADDR_PORT, oldval1); diff --git a/sys/i386/pci/pci_pir.c b/sys/i386/pci/pci_pir.c index 91746fe..0e3b8c3 100644 --- a/sys/i386/pci/pci_pir.c +++ b/sys/i386/pci/pci_pir.c @@ -732,7 +732,7 @@ pcireg_cfgopen(void) devmax = 32; outl(CONF1_ADDR_PORT, CONF1_ENABLE_CHK); - outb(CONF1_ADDR_PORT + 3, 0); + DELAY(1); mode1res = inl(CONF1_ADDR_PORT); outl(CONF1_ADDR_PORT, oldval1); |
