[Bug 260274] bnxt(4) eats incoming gre packets passing through unless promisc is toggled on

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=260274

            Bug ID: 260274
           Summary: bnxt(4) eats incoming gre packets passing through
                    unless promisc is toggled on
           Product: Base System
           Version: CURRENT
          Hardware: amd64
                OS: Any
            Status: New
          Severity: Affects Only Me
          Priority: ---
         Component: kern
          Assignee: bugs@FreeBSD.org
          Reporter: knan-bfo@modirum.com

On our routers with BSDRP 1.99, we have a weird problem with bnxt(4)
interfaces. GRE packets coming in on the bnxt interfaces are never seen and
forwarded by the router, unless we toggle promisc mode on.

If we bring the interface up in promisc mode after a reboot, the packets are
also eaten and not visible in tcpdump. If we toggle promisc off and on again
with ifconfig, gre packets are suddenly visible and passed on by routing.

We have not set any bnxt-specific sysctls.

BSDRP 1.99 uses FreeBSD 14-head 7b8696bf128

from dmesg:

bnxt0: <Broadcom BCM57416 NetXtreme-E 10GBase-T Ethernet> mem
0x18020110000-0x1802011ffff,0x18020000000-0x180200fffff,0x18020122000-0x18020123fff
at device 0.0 on pci6
bnxt0: Using 256 TX descriptors and 256 RX descriptors
bnxt0: Using 16 RX queues 16 TX queues
bnxt0: Using MSI-X interrupts with 17 vectors
bnxt0: Ethernet address: <snip>
bnxt0: netmap queues/slots: TX 16/256, RX 16/256
bnxt1: <Broadcom BCM57416 NetXtreme-E 10GBase-T Ethernet> mem
0x18020100000-0x1802010ffff,0x1801ff00000-0x1801fffffff,0x18020120000-0x18020121fff
at device 0.1 on pci6
bnxt1: Using 256 TX descriptors and 256 RX descriptors
bnxt1: Using 16 RX queues 16 TX queues
bnxt1: Using MSI-X interrupts with 17 vectors
bnxt1: Ethernet address: <snip>
bnxt1: netmap queues/slots: TX 16/256, RX 16/256

versions:
dev.bnxt.1.ver.hwrm_min_ver: 1.2.2
dev.bnxt.1.ver.package_ver: <unknown>
dev.bnxt.1.ver.chip_type: ASIC
dev.bnxt.1.ver.chip_bond_id: 0
dev.bnxt.1.ver.chip_metal: 1
dev.bnxt.1.ver.chip_rev: 1
dev.bnxt.1.ver.chip_num: 5848
dev.bnxt.1.ver.phy_partnumber: 
dev.bnxt.1.ver.phy_vendor: 
dev.bnxt.1.ver.roce_fw_name: BONO_FW
dev.bnxt.1.ver.netctrl_fw_name: KONG_FW
dev.bnxt.1.ver.mgmt_fw_name: AFW_214.4.28.0
dev.bnxt.1.ver.hwrm_fw_name: CHIMP_FW
dev.bnxt.1.ver.phy: 13.1.11
dev.bnxt.1.ver.roce_fw: 214.0.194
dev.bnxt.1.ver.netctrl_fw: 214.0.255
dev.bnxt.1.ver.mgmt_fw: 214.4.28
dev.bnxt.1.ver.hwrm_fw: 214.4.65
dev.bnxt.1.ver.driver_hwrm_if: 1.8.1.7
dev.bnxt.1.ver.hwrm_if: 1.10.1

-- 
You are receiving this mail because:
You are the assignee for the bug.