From mboxrd@z Thu Jan 1 00:00:00 1970 References: <5D4944CC.30100@huawei.com> <20190806094259.GC3066@work-vm> From: piaojun Message-ID: <5D4963F5.7030407@huawei.com> Date: Tue, 6 Aug 2019 19:26:45 +0800 MIME-Version: 1.0 In-Reply-To: <20190806094259.GC3066@work-vm> Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 7bit Subject: Re: [Virtio-fs] [QUESTION] Can not enable dax option List-Id: Development discussions about virtio-fs List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Dr. David Alan Gilbert" Cc: virtio-fs@redhat.com Hi Dave, On 2019/8/6 17:42, Dr. David Alan Gilbert wrote: > * piaojun (piaojun@huawei.com) wrote: >> Hi, >> >> Dax could not be enabled by run the following command: >> >> QEMU command: >> qemu-system-x86_64 -M pc -cpu host --enable-kvm -smp 2 \ >> -m 4G,maxmem=4G -object memory-backend-file,id=mem,size=4G,mem-path=/dev/shm,share=on -numa node,memdev=mem \ >> -drive if=none,id=root,format=qcow2,file=centos_7.3_64.qcow2 -device virtio-scsi -device scsi-disk,drive=root,bootindex=1 -object iothread,id=io \ >> -device virtio-scsi-pci,iothread=io -net nic,model=virtio -vnc 0.0.0.0:0 \ >> -chardev socket,id=char0,path=/tmp/vhostqemu -device vhost-user-fs-pci,queue-size=1024,chardev=char0,tag=myfs,cache-size=2G >> >> Guest command: >> # mount -t virtio_fs myfs /mnt/virtiofs -o rootmode=040000,user_id=0,group_id=0,dax >> >> When I rmmod/insmod again, the error message shows that something is >> wrong with 'cache capability': >> >> virtio_pci_find_shm_cap: shm cap with bad size offset: 164 size: 24 >> virtio_fs virtio2: virtio_fs_setup_dax: No cache capability > > That's odd; because '24' seems the right size for a 'struct > virtio_pci_cap64' and is what I get on my machine. > > Can you add: > > diff --git a/drivers/virtio/virtio_pci_modern.c b/drivers/virtio/virtio_pci_modern.c > index fcdabe85d82d..34e89474a8b7 100644 > --- a/drivers/virtio/virtio_pci_modern.c > +++ b/drivers/virtio/virtio_pci_modern.c > @@ -473,6 +473,7 @@ static int virtio_pci_find_shm_cap(struct pci_dev *dev, > __func__, pos, cap_len); > continue; > } > + printk(KERN_INFO "%s: cap_len should be: %d\n", __func__, sizeof(struct virtio_pci_cap64)); > > pci_read_config_byte(dev, pos + offsetof(struct virtio_pci_cap, > id), > > and tell us what it says? > > Just check that your kernel's include/uapi/linux/virtio_pci.h and > daemon's include/standard-headers/linux/virtio_pci.h both match > and show the same definition for virtio_pci_cap64 > > Dave Thanks for your help, and I solved this problem by reinstalling Guest kernel: # make modules_install -j 16 # make install -j 16 1. The virtio_pci.h has the same definition for virtio_pci_cap64 in both kernel and daemon; 2. As you said, the reason is probably that Guest kernel still used virtio_pci_cap which not matches virtio_pci_cap64. I guess the Guest still used the old virtio driver which cause the problem. Thanks, Jun > > >> >> Code url: >> https://github.com/rhvgoyal/linux/commits/virtio-fs-dev-5.1 >> https://gitlab.com/virtio-fs/qemu/tree/virtio-fs-dev >> >> Thanks, >> Jun >> >> _______________________________________________ >> Virtio-fs mailing list >> Virtio-fs@redhat.com >> https://www.redhat.com/mailman/listinfo/virtio-fs > -- > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK > . >