DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [Bug 585] Direct PCI access violation with VFIO and lockdown mode
@ 2020-11-30 14:41 bugzilla
  2020-12-17 16:57 ` bugzilla
  0 siblings, 1 reply; 2+ messages in thread
From: bugzilla @ 2020-11-30 14:41 UTC (permalink / raw)
  To: dev

https://bugs.dpdk.org/show_bug.cgi?id=585

            Bug ID: 585
           Summary: Direct PCI access violation with VFIO and lockdown
                    mode
           Product: DPDK
           Version: 19.11
          Hardware: ARM
                OS: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: Normal
         Component: core
          Assignee: dev@dpdk.org
          Reporter: luca.boccassi@gmail.com
  Target Milestone: ---

With DPDK 19.11 on an arm64 platform using vfio-pci, we are getting errors on
initialization when kernel lockdown mode is enabled. Lockdown restricts direct
PCI access, but intentionally allows vfio as it's mediated access. On x86 this
works fine, no illegal accesses are reported, and it seems to be a problem only
on arm64.

[    2.687022] vfio-pci 0008:01:00.1: vfio_ecap_init: hiding ecap 0x0@0x250     
[    3.391535] vfio-pci 0000:01:00.1: enabling device (0000 -> 0002)            
[    3.507258] Lockdown: dpdk-app: direct PCI access is restricted; see man
kernel_lockdown.7                                                           
[    3.507263] ------------[ cut here ]------------                             
[    3.507264] pci_write_config:
/devices/platform/48000000.pcie/pci0000:00/0000:00:00.0/0000:01:00.1            
[    3.507303] WARNING: CPU: 4 PID: 1473 at /drivers/pci/pci-sysfs.c:765
pci_write_config+0x1bc/0x1c4                                                    
[    3.507304] Modules linked in: bnxt_en pcie_iproc_platform pcie_iproc        
[    3.507312] CPU: 4 PID: 1473 Comm: dpdk-app Not tainted 5.4.72 #1            
[    3.507313] Hardware name: Overlake (DT)                                     
[    3.507315] pstate: 60400005 (nZCv daif +PAN -UAO)                           
[    3.507316] pc : pci_write_config+0x1bc/0x1c4                                
[    3.507318] lr : pci_write_config+0x1bc/0x1c4                                
[    3.507318] sp : ffff8000113ebcb0                                            
[    3.507320] x29: ffff8000113ebcf0 x28: ffff0008e3758000                      
[    3.507321] x27: 0000000000000000 x26: 0000000000000000                      
[    3.507323] x25: 0000000056000000 x24: ffffd539ba59a000                      
[    3.507325] x23: ffff0008e5ad90b0 x22: 0000000000000002                      
[    3.507326] x21: 00000000ffffffff x20: 0000000000000004                      
[    3.507327] x19: ffff0008d8594a00 x18: 0000000000000000                      
[    3.507329] x17: 0000000000000078 x16: 0000000000015f24                      
[    3.507330] x15: 0000000000000022 x14: 0000000000000068                      
[    3.507332] x13: 000000000000a144 x12: 0000000000000000                      
[    3.507333] x11: 0000000000000001 x10: 0000000000000007                      
[    3.507334] x9 : 31c97f42d035aa00 x8 : 31c97f42d035aa00                      
[    3.507336] x7 : 0000000000000000 x6 : ffffd539ba5aa81e                      
[    3.507337] x5 : 0000000000000000 x4 : 0000000000000004                      
[    3.507339] x3 : 000000000000312e x2 : fffffffffffffe24                      
[    3.507340] x1 : 0000000000000000 x0 : 0000000000000056                      
[    3.507342] Call trace:                                                      
[    3.507343]  pci_write_config+0x1bc/0x1c4                                    
[    3.507351]  sysfs_kf_bin_write+0x6c/0x94
[    3.507353]  kernfs_fop_write+0x1d4/0x2b0
[    3.507355]  __vfs_write+0x54/0x198
[    3.507356]  vfs_write+0xcc/0x18c
[    3.507358]  __arm64_sys_pwrite64+0x80/0xc0
[    3.507362]  el0_svc_handler+0xe0/0x140
[    3.507364]  el0_svc+0x8/0x208
[    3.507365] ---[ end trace 116c21626a554aca ]---

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

^ permalink raw reply	[flat|nested] 2+ messages in thread

* [dpdk-dev] [Bug 585] Direct PCI access violation with VFIO and lockdown mode
  2020-11-30 14:41 [dpdk-dev] [Bug 585] Direct PCI access violation with VFIO and lockdown mode bugzilla
@ 2020-12-17 16:57 ` bugzilla
  0 siblings, 0 replies; 2+ messages in thread
From: bugzilla @ 2020-12-17 16:57 UTC (permalink / raw)
  To: dev

https://bugs.dpdk.org/show_bug.cgi?id=585

Luca Boccassi (luca.boccassi@gmail.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |INVALID
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #8 from Luca Boccassi (luca.boccassi@gmail.com) ---
So Stephen checked all that, and it turns out, while the accesses are coming
from the DPDK application, they are actually from a vendor library, not DPDK.
So closing as invalid. Thanks for the tips!

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

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2020-12-17 16:57 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-11-30 14:41 [dpdk-dev] [Bug 585] Direct PCI access violation with VFIO and lockdown mode bugzilla
2020-12-17 16:57 ` bugzilla

DPDK patches and discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://inbox.dpdk.org/dev/0 dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 dev dev/ https://inbox.dpdk.org/dev \
		dev@dpdk.org
	public-inbox-index dev

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git