[PATCH 2/4]bnx2x: PCI configuration bug on big-endian
From: Eilon Greenstein <hidden>
Date: 2008-11-03 16:11:32
Subsystem:
networking drivers, the rest · Maintainers:
Andrew Lunn, "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Linus Torvalds
The current code read nothing but zeros on big-endian (wrong part of the 32bits). This caused poor performance on big-endian machines. Though this issue did not cause the system to crash, the performance is significantly better with the fix so I view it as critical bug fix. Signed-off-by: Eilon Greenstein <redacted> --- drivers/net/bnx2x_init.h | 9 +++++---- 1 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/net/bnx2x_init.h b/drivers/net/bnx2x_init.h
index 130927c..a6c0b3a 100644
--- a/drivers/net/bnx2x_init.h
+++ b/drivers/net/bnx2x_init.h@@ -564,14 +564,15 @@ static const struct arb_line write_arb_addr[NUM_WR_Q-1] = { static void bnx2x_init_pxp(struct bnx2x *bp) { + u16 devctl; int r_order, w_order; u32 val, i; pci_read_config_word(bp->pdev, - bp->pcie_cap + PCI_EXP_DEVCTL, (u16 *)&val); - DP(NETIF_MSG_HW, "read 0x%x from devctl\n", (u16)val); - w_order = ((val & PCI_EXP_DEVCTL_PAYLOAD) >> 5); - r_order = ((val & PCI_EXP_DEVCTL_READRQ) >> 12); + bp->pcie_cap + PCI_EXP_DEVCTL, &devctl); + DP(NETIF_MSG_HW, "read 0x%x from devctl\n", devctl); + w_order = ((devctl & PCI_EXP_DEVCTL_PAYLOAD) >> 5); + r_order = ((devctl & PCI_EXP_DEVCTL_READRQ) >> 12); if (r_order > MAX_RD_ORD) { DP(NETIF_MSG_HW, "read order of %d order adjusted to %d\n",
--
1.5.4.3