| Message ID | 20210614174454.903555-2-vgoyal@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Add support to boot virtiofs and 9pfs as rootfs | expand |
On Mon, 14 Jun 2021 at 19:45, Vivek Goyal <vgoyal@redhat.com> wrote: > > We want to be able to mount virtiofs as rootfs and pass appropriate > kernel command line. Right now there does not seem to be a good way > to do that. If I specify "root=myfs rootfstype=virtiofs", system > panics. > > virtio-fs: tag </dev/root> not found > .. > .. > [ end Kernel panic - not syncing: VFS: Unable to mount root fs on > +unknown-block(0,0) ] > > Basic problem here is that kernel assumes that device identifier > passed in "root=" is a block device. But there are few execptions > to this rule to take care of the needs of mtd, ubi, NFS and CIFS. > > For example, mtd and ubi prefix "mtd:" or "ubi:" respectively. > > "root=mtd:<identifier>" or "root=ubi:<identifier>" > > NFS and CIFS use "root=/dev/nfs" and CIFS passes "root=/dev/cifs" and > actual root device details come from filesystem specific kernel > command line options. > > virtiofs does not seem to fit in any of the above categories. In fact > we have 9pfs which can be used to boot from but it also does not > have a proper syntax to specify rootfs and does not fit into any of > the existing syntax. They both expect a device "tag" to be passed > in a device to be mounted. And filesystem knows how to parse and > use "tag". > > So there seem to be a class of filesystems which specify root device > using a "tag" which is understood by the filesystem. And filesystem > simply expects that "tag" to be passed as "source" of mount and > how to mount filesystem using that "tag" is responsibility of filesystem. > > This patch proposes that we internally create a list of filesystems > which pass a "tag" in "root=<tag>" and expect that tag to be passed > as "source" of mount. With this patch I can boot into virtiofs rootfs > with following syntax. > > "root=myfs rootfstype=virtiofs rw" The syntax and the implementation looks good. Acked-by: Miklos Szeredi <mszeredi@redhat.com>
On Fri, Jun 18, 2021 at 09:07:14AM +0200, Miklos Szeredi wrote: > On Mon, 14 Jun 2021 at 19:45, Vivek Goyal <vgoyal@redhat.com> wrote: > > > > We want to be able to mount virtiofs as rootfs and pass appropriate > > kernel command line. Right now there does not seem to be a good way > > to do that. If I specify "root=myfs rootfstype=virtiofs", system > > panics. > > > > virtio-fs: tag </dev/root> not found > > .. > > .. > > [ end Kernel panic - not syncing: VFS: Unable to mount root fs on > > +unknown-block(0,0) ] > > > > Basic problem here is that kernel assumes that device identifier > > passed in "root=" is a block device. But there are few execptions > > to this rule to take care of the needs of mtd, ubi, NFS and CIFS. > > > > For example, mtd and ubi prefix "mtd:" or "ubi:" respectively. > > > > "root=mtd:<identifier>" or "root=ubi:<identifier>" > > > > NFS and CIFS use "root=/dev/nfs" and CIFS passes "root=/dev/cifs" and > > actual root device details come from filesystem specific kernel > > command line options. > > > > virtiofs does not seem to fit in any of the above categories. In fact > > we have 9pfs which can be used to boot from but it also does not > > have a proper syntax to specify rootfs and does not fit into any of > > the existing syntax. They both expect a device "tag" to be passed > > in a device to be mounted. And filesystem knows how to parse and > > use "tag". > > > > So there seem to be a class of filesystems which specify root device > > using a "tag" which is understood by the filesystem. And filesystem > > simply expects that "tag" to be passed as "source" of mount and > > how to mount filesystem using that "tag" is responsibility of filesystem. > > > > This patch proposes that we internally create a list of filesystems > > which pass a "tag" in "root=<tag>" and expect that tag to be passed > > as "source" of mount. With this patch I can boot into virtiofs rootfs > > with following syntax. > > > > "root=myfs rootfstype=virtiofs rw" > > The syntax and the implementation looks good. > > Acked-by: Miklos Szeredi <mszeredi@redhat.com> Thanks Miklos. Now Christoph has come up with another patch series to achieve the same. And that patches series looks better than my patches. https://lore.kernel.org/linux-fsdevel/20210617153649.1886693-1-hch@lst.de/ There are couple of minor issues to be taken care of. I am hoping after that these patches can be merged. Thanks Vivek
diff --git a/init/do_mounts.c b/init/do_mounts.c index a78e44ee6adb..2a18238f4962 100644 --- a/init/do_mounts.c +++ b/init/do_mounts.c @@ -31,6 +31,12 @@ int root_mountflags = MS_RDONLY | MS_SILENT; static char * __initdata root_device_name; static char __initdata saved_root_name[64]; +static char *__initdata tag_based_rootfs[] = { +#if IS_BUILTIN(CONFIG_VIRTIO_FS) + "virtiofs", +#endif +}; +static bool __initdata tag_based_root; static int root_wait; dev_t ROOT_DEV; @@ -552,6 +558,14 @@ void __init mount_root(void) return; } #endif + if (tag_based_root) { + if (!do_mount_root(root_device_name, root_fs_names, + root_mountflags, root_mount_data)) + return; + panic("VFS: Unable to mount root \"%s\" via \"%s\"\n", + root_device_name, root_fs_names); + } + #ifdef CONFIG_BLOCK { int err = create_dev("/dev/root", ROOT_DEV); @@ -563,6 +577,20 @@ void __init mount_root(void) #endif } +static bool is_tag_based_rootfs(char *name) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(tag_based_rootfs); i++) { + int name_len = strlen(tag_based_rootfs[i]) + 1; + + if (!strncmp(tag_based_rootfs[i], name, name_len)) + return true; + } + + return false; +} + /* * Prepare the namespace - decide what/where to mount, load ramdisks, etc. */ @@ -593,6 +621,10 @@ void __init prepare_namespace(void) goto out; } ROOT_DEV = name_to_dev_t(root_device_name); + if (ROOT_DEV == 0 && root_fs_names) { + if (is_tag_based_rootfs(root_fs_names)) + tag_based_root = true; + } if (strncmp(root_device_name, "/dev/", 5) == 0) root_device_name += 5; }
We want to be able to mount virtiofs as rootfs and pass appropriate kernel command line. Right now there does not seem to be a good way to do that. If I specify "root=myfs rootfstype=virtiofs", system panics. virtio-fs: tag </dev/root> not found .. .. [ end Kernel panic - not syncing: VFS: Unable to mount root fs on +unknown-block(0,0) ] Basic problem here is that kernel assumes that device identifier passed in "root=" is a block device. But there are few execptions to this rule to take care of the needs of mtd, ubi, NFS and CIFS. For example, mtd and ubi prefix "mtd:" or "ubi:" respectively. "root=mtd:<identifier>" or "root=ubi:<identifier>" NFS and CIFS use "root=/dev/nfs" and CIFS passes "root=/dev/cifs" and actual root device details come from filesystem specific kernel command line options. virtiofs does not seem to fit in any of the above categories. In fact we have 9pfs which can be used to boot from but it also does not have a proper syntax to specify rootfs and does not fit into any of the existing syntax. They both expect a device "tag" to be passed in a device to be mounted. And filesystem knows how to parse and use "tag". So there seem to be a class of filesystems which specify root device using a "tag" which is understood by the filesystem. And filesystem simply expects that "tag" to be passed as "source" of mount and how to mount filesystem using that "tag" is responsibility of filesystem. This patch proposes that we internally create a list of filesystems which pass a "tag" in "root=<tag>" and expect that tag to be passed as "source" of mount. With this patch I can boot into virtiofs rootfs with following syntax. "root=myfs rootfstype=virtiofs rw" This patch adds support for virtiofs and next patch adds 9p to the list. And any other file system which is fine with these semantics, should be able to work with it easily. Signed-off-by: Vivek Goyal <vgoyal@redhat.com> --- init/do_mounts.c | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+)