Hi Vipin, Please see inline.
> -----Original Message----- > From: dev <dev-boun...@dpdk.org> On Behalf Of Varghese, Vipin > Sent: Wednesday, April 1, 2020 8:45 AM > To: Vamsi Krishna Attunuru <vattun...@marvell.com>; dev@dpdk.org > Cc: Jerin Jacob Kollanukkaran <jer...@marvell.com>; > alex.william...@redhat.com; tho...@monjalon.net; > david.march...@redhat.com > Subject: Re: [dpdk-dev] [RFC v1 1/1] vfio: set vf token and gain vf device > access > > Snipped > > diff --git a/lib/librte_eal/linux/eal/eal_vfio.c > > b/lib/librte_eal/linux/eal/eal_vfio.c > > index 01b5ef3..e2fdd35 100644 > > --- a/lib/librte_eal/linux/eal/eal_vfio.c > > +++ b/lib/librte_eal/linux/eal/eal_vfio.c > > DPDK supports freebsd where `/dev/pci` can be probed like Linux > `/sys/bus/pci`. Will you be adding the functions in > `lib/librte_eal/freebsd/eal/` folder or implement as `NOT supported`? > > What about window ` lib/librte_eal/windows/eal/`? If other kernel (freebsd & windows) drivers also support this vf token scheme, functions can be added in respective eal implementations. > > snipped > > + snprintf(linkname, sizeof(linkname), > > + "%s/%s/physfn", sysfs_base, dev_addr); > > Would you like to check the pointer sysfs_base or dev_addr? Physfn pointer is being checked here. > > > int > > rte_vfio_setup_device(const char *sysfs_base, const char *dev_addr, > > int *vfio_dev_fd, struct vfio_device_info *device_info) @@ - > > Snipped > > > + ret = is_vf_token_required(sysfs_base, dev_addr); > > The return value from the above function is 0 or -1. Function also returns 1. > > > + /* if negative, something failed */ > > + if (ret < 0) > > + return -1; > > + > > + if (ret == 0) { > > Will it be ok to skip this check? No, it's a valid case. > > > + /* vf_token required to open device file descriptor */ > > + rte_uuid_unparse(uuid_token, > > + vf_token, sizeof(vf_token)); > > + snprintf(dev, sizeof(dev), > > + "%s vf_token=%s", dev_addr, vf_token); > > + } > > snipped > > general comment `can we use EAL LOG for info, warn, error`?