Workaround to prevent VMware from melting down. The pseudo PCnet interface

in VMware reports 0x00000000 in the PCI subsystem ID register, but
0x10001000 when you read the mirror registers in I/O space. This causes
pcn_probe() to think it's found a card in 32-bit mode, and performing
a 32-bit I/O access makes on a 16-bit port makes VMware go boom. Special
case the 0x10001000 value until somebody at VMware grows a clue.

Finally discovered by: Andrew Gallatin

1 files changed, 9 insertions, 1 deletions

diff --git a/sys/pci/if_pcn.c b/sys/pci/if_pcn.c
index eced63224707..6bf15e0c7ead 100644
--- a/sys/pci/if_pcn.c
+++ b/sys/pci/if_pcn.c
@@ -440,8 +440,16 @@ static int pcn_probe(dev)
 			chip_id = pcn_bcr_read16(sc, PCN_BCR_PCISUBSYSID);
 			chip_id <<= 16;
 			chip_id |= pcn_bcr_read16(sc, PCN_BCR_PCISUBVENID);
+			/*
+			 * Note III: the test for 0x10001000 is a hack to
+			 * pacify VMware, who's pseudo-PCnet interface is
+			 * broken. Reading the subsystem register from PCI
+			 * config space yeilds 0x00000000 while reading the
+			 * same value from I/O space yeilds 0x10001000. It's
+			 * not supposed to be that way.
+			 */
 			if (chip_id == pci_read_config(dev,
-			    PCIR_SUBVEND_0, 4)) {
+			    PCIR_SUBVEND_0, 4) || chip_id == 0x10001000) {
 				/* We're in 16-bit mode. */
 				chip_id = pcn_csr_read16(sc, PCN_CSR_CHIPID1);
 				chip_id <<= 16;