https://bugs.dpdk.org/show_bug.cgi?id=649
Bug ID: 649
Summary: [dpdk-21.05] external_memory/test_VFIO_PCI_xmem: start
testpmd failed
Product: DPDK
Version: unspecified
Hardware: x86
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: Normal
Component: testpmd
Assignee: dev@dpdk.org
Reporter: linglix.c...@intel.com
Target Milestone: ---
With the lastest version on dpdk.org (i.e. commit id
44db5a5cf23f6157448fdb692efab605544762f3)
Reproduce Step:
1.bind nic to vfio-pci
usertools/dpdk-devbind.py --force --bind=vfio-pci 0000:83:00.0 0000:83:00.1
0000:83:00.2 0000:83:00.3
2. start pf testpmd, but it failed
./x86_64-native-linuxapp-gcc/app/dpdk-testpmd -l 1,2,3,4 -n 4
--file-prefix=dpdk_1766_20210302180024 – --mp-alloc=xmem -i
EAL: Detected 88 lcore(s)
EAL: Detected 2 NUMA nodes
EAL: Detected static linkage of DPDK
EAL: Multi-process socket /var/run/dpdk/dpdk_1766_20210302180024/mp_socket
EAL: Selected IOVA mode 'VA'
EAL: 1024 hugepages of size 2097152 reserved, but no mounted hugetlbfs found
for that size
EAL: Probing VFIO support...
EAL: VFIO support initialized
EAL: using IOMMU type 1 (Type 1)
EAL: Probe PCI driver: net_e1000_igb (8086:1521) device: 0000:83:00.0 (socket
1)
EAL: Probe PCI driver: net_e1000_igb (8086:1521) device: 0000:83:00.1 (socket
1)
EAL: Probe PCI driver: net_e1000_igb (8086:1521) device: 0000:83:00.2 (socket
1)
EAL: Probe PCI driver: net_e1000_igb (8086:1521) device: 0000:83:00.3 (socket
1)
EAL: No legacy callbacks, legacy socket not created
Interactive-mode selected
testpmd: create a new mbuf pool <mb_pool_0>: n=171456, size=2176, socket=0
EAL: WARNING! Base virtual address hint (0x2200005000 Unable to render embedded
object: File (= 0x7f468969c000) not respected) not found.
EAL: This may cause issues with mapping memory into secondary processes
EAL: cannot set up DMA remapping, error 28 (No space left on device)
EAL: cannot set up DMA remapping, error 28 (No space left on device)
EAL: cannot set up DMA remapping, error 28 (No space left on device)
Expect results: no error log and start PF testpmd normal.
Is this issue a regression: Y
Version the regression was introduced: Specify git id if known.
First bad commit:
commit 016763c219580292c8b05059c7452a7a11d0d19e
Author: Nithin Dabilpuram <ndabilpu...@marvell.com>
Date: Fri Jan 15 13:02:41 2021 +0530
vfio: do not merge contiguous areas
In order to save DMA entries limited by kernel both for external
memory and hugepage memory, an attempt was made to map physically
contiguous memory in one go. This cannot be done as VFIO IOMMU type1
does not support partially unmapping a previously mapped memory
region while Heap can request for multi page mapping and
partial unmapping.
Hence for going back to old method of mapping/unmapping at
memseg granularity, this commit reverts
commit d1c7c0cdf7ba ("vfio: map contiguous areas in one go")
Also add documentation on what module parameter needs to be used
to increase the per-container dma map limit for VFIO.
Fixes: d1c7c0cdf7ba ("vfio: map contiguous areas in one go")
Cc: sta...@dpdk.org
Signed-off-by: Nithin Dabilpuram <ndabilpu...@marvell.com>
Acked-by: Anatoly Burakov <anatoly.bura...@intel.com>
Acked-by: David Christensen <d...@linux.vnet.ibm.com>
--
You are receiving this mail because:
You are the assignee for the bug.
