Re: [REGRESSION] mlx5: Driver remove during hot unplug is broken

2020-07-10 Thread Niklas Schnelle
On 7/9/20 8:34 PM, Parav Pandit wrote: > On 7/9/2020 3:36 PM, Niklas Schnelle wrote: >> >> On 7/8/20 5:44 PM, Parav Pandit wrote: ... snip .. > >>> >> As is the patch above fixes the dereference but results in the same >> completion error >&g

Re: [REGRESSION] mlx5: Driver remove during hot unplug is broken

2020-07-09 Thread Niklas Schnelle
On 7/8/20 5:44 PM, Parav Pandit wrote: ... snip .. >> > > It is likely because events_cleanup() freed the memory using kvfree() that > health recovery context is trying to access in notifier chain. > > While reviewing I see few more errors as below. > (a) mlx5_pci_err_detected() invokes mlx5_d

Re: [REGRESSION] mlx5: Driver remove during hot unplug is broken

2020-07-08 Thread Niklas Schnelle
Hi Parav, Hi Shay, On 7/8/20 12:43 PM, Parav Pandit wrote: > Hi Niklas, > ... snip ... >>> > > Sorry for my late response. > Yes, this looks good and I also found same in my analysis. > With latest code mlx5_pci_close() already does drain_health_wq(), so the > additional call in remove_one() is

mlx5 hot unplug regression on z/VM

2020-07-06 Thread Niklas Schnelle
mlx5_core driver to make more progress. Best regards, Niklas Schnelle [ 70.773478] IPv6: ADDRCONF(NETDEV_CHANGE): ens514np0: link becomes ready [ 70.858688] RPC: Registered named UNIX socket transport module. [ 70.858709] RPC: Registered udp transport module. [ 70.858714] RPC: Registered tcp t

Re: [REGRESSION] mlx5: Driver remove during hot unplug is broken

2020-06-15 Thread Niklas Schnelle
Hello Saeed, On 6/13/20 12:01 AM, Saeed Mahameed wrote: > On Fri, 2020-06-12 at 15:09 +0200, Niklas Schnelle wrote: >> Hello Parav, Hello Saeed, >> ... snip ... >> >> So without really knowing anything about these functions I would >> guess that with the de

[REGRESSION] mlx5: Driver remove during hot unplug is broken

2020-06-12 Thread Niklas Schnelle
es that sound plausible to you? Best regards, Niklas Schnelle [0] dmesg output: [ 36.447442] mlx5_core :00:00.0: poll_health:694:(pid 0): Fatal error 1 detected [ 36.447450] mlx5_core :00:00.0: print_health_info:372:(pid 0): assert_var[0] 0x [ 36.447453] mlx5_core :

Re: [PATCH 1/1] net/mlx5: Call pci_disable_sriov() on remove

2020-04-30 Thread Niklas Schnelle
On 4/30/20 9:47 PM, Niklas Schnelle wrote: > > > On 4/30/20 5:58 PM, Saeed Mahameed wrote: >> On Thu, 2020-04-30 at 14:03 +0200, Niklas Schnelle wrote: >>> as described in Documentation/PCI/pci-iov-howto.rst a driver with SR- >>> IOV >>> support sho

Re: [PATCH 0/1] net/mlx5: Call pci_disable_sriov() on remove

2020-04-30 Thread Niklas Schnelle
On 4/30/20 6:13 PM, Saeed Mahameed wrote: > On Thu, 2020-04-30 at 14:03 +0200, Niklas Schnelle wrote: >> Hello, >> >> I'm currently working on improvements in PF-VF handling on s390. One >> thing that >> may be a bit special for us is that the s390 hotpl

Re: [PATCH 1/1] net/mlx5: Call pci_disable_sriov() on remove

2020-04-30 Thread Niklas Schnelle
On 4/30/20 5:58 PM, Saeed Mahameed wrote: > On Thu, 2020-04-30 at 14:03 +0200, Niklas Schnelle wrote: >> as described in Documentation/PCI/pci-iov-howto.rst a driver with SR- >> IOV >> support should call pci_disable_sriov() in the remove handler. > > Hi

[PATCH 1/1] net/mlx5: Call pci_disable_sriov() on remove

2020-04-30 Thread Niklas Schnelle
mlx5_device_disable_sriov() which is called by both. Signed-off-by: Niklas Schnelle --- drivers/net/ethernet/mellanox/mlx5/core/sriov.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c index

[PATCH 0/1] net/mlx5: Call pci_disable_sriov() on remove

2020-04-30 Thread Niklas Schnelle
ly missing something here in that case excuse my ignorance. Best regards, Niklas Schnelle Niklas Schnelle (1): net/mlx5: Call pci_disable_sriov() on remove drivers/net/ethernet/mellanox/mlx5/core/sriov.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) -- 2.17.1