| Message ID | 20210811122811.2288041-1-yangerkun@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | ramfs: fix mount source show for ramfs | expand |
Hi, this patch seems still leave in linux-next, should we pull it to mainline? 在 2021/8/11 20:28, yangerkun 写道: > ramfs_parse_param does not parse key "source", and will convert > -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in > vfs_parse_fs_param, which lead always "none" mount source for ramfs. Fix > it by parse "source" in ramfs_parse_param. > > Signed-off-by: yangerkun <yangerkun@huawei.com> > --- > fs/ramfs/inode.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c > index 65e7e56005b8..0d7f5f655fd8 100644 > --- a/fs/ramfs/inode.c > +++ b/fs/ramfs/inode.c > @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context *fc, struct fs_parameter *param) > struct ramfs_fs_info *fsi = fc->s_fs_info; > int opt; > > + opt = vfs_parse_fs_param_source(fc, param); > + if (opt != -ENOPARAM) > + return opt; > + > opt = fs_parse(fc, ramfs_fs_parameters, param, &result); > if (opt < 0) { > /* >
On Wed, 8 Sep 2021 16:56:25 +0800 yangerkun <yangerkun@huawei.com> wrote: > 在 2021/8/11 20:28, yangerkun 写道: > > ramfs_parse_param does not parse key "source", and will convert > > -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in > > vfs_parse_fs_param, which lead always "none" mount source for ramfs. Fix > > it by parse "source" in ramfs_parse_param. > > > > Signed-off-by: yangerkun <yangerkun@huawei.com> > > --- > > fs/ramfs/inode.c | 4 ++++ > > 1 file changed, 4 insertions(+) > > > > diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c > > index 65e7e56005b8..0d7f5f655fd8 100644 > > --- a/fs/ramfs/inode.c > > +++ b/fs/ramfs/inode.c > > @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context *fc, struct fs_parameter *param) > > struct ramfs_fs_info *fsi = fc->s_fs_info; > > int opt; > > > > + opt = vfs_parse_fs_param_source(fc, param); > > + if (opt != -ENOPARAM) > > + return opt; > > + > > opt = fs_parse(fc, ramfs_fs_parameters, param, &result); > > if (opt < 0) { > > /* > > (top-posting repaired) > Hi, this patch seems still leave in linux-next, should we pull it to > mainline? I was hoping for a comment from Al?
在 2021/9/9 6:39, Andrew Morton 写道: > On Wed, 8 Sep 2021 16:56:25 +0800 yangerkun <yangerkun@huawei.com> wrote: > >> 在 2021/8/11 20:28, yangerkun 写道: >>> ramfs_parse_param does not parse key "source", and will convert >>> -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in >>> vfs_parse_fs_param, which lead always "none" mount source for ramfs. Fix >>> it by parse "source" in ramfs_parse_param. >>> >>> Signed-off-by: yangerkun <yangerkun@huawei.com> >>> --- >>> fs/ramfs/inode.c | 4 ++++ >>> 1 file changed, 4 insertions(+) >>> >>> diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c >>> index 65e7e56005b8..0d7f5f655fd8 100644 >>> --- a/fs/ramfs/inode.c >>> +++ b/fs/ramfs/inode.c >>> @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context *fc, struct fs_parameter *param) >>> struct ramfs_fs_info *fsi = fc->s_fs_info; >>> int opt; >>> >>> + opt = vfs_parse_fs_param_source(fc, param); >>> + if (opt != -ENOPARAM) >>> + return opt; >>> + >>> opt = fs_parse(fc, ramfs_fs_parameters, param, &result); >>> if (opt < 0) { >>> /* >>> > > (top-posting repaired) > >> Hi, this patch seems still leave in linux-next, should we pull it to >> mainline? > > I was hoping for a comment from Al? Hi, Al, Can you help to review this patch... Thanks, Kun. > . >
在 2021/9/9 16:37, yangerkun 写道: > > > 在 2021/9/9 6:39, Andrew Morton 写道: >> On Wed, 8 Sep 2021 16:56:25 +0800 yangerkun <yangerkun@huawei.com> wrote: >> >>> 在 2021/8/11 20:28, yangerkun 写道: >>>> ramfs_parse_param does not parse key "source", and will convert >>>> -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in >>>> vfs_parse_fs_param, which lead always "none" mount source for ramfs. >>>> Fix >>>> it by parse "source" in ramfs_parse_param. >>>> >>>> Signed-off-by: yangerkun <yangerkun@huawei.com> >>>> --- >>>> fs/ramfs/inode.c | 4 ++++ >>>> 1 file changed, 4 insertions(+) >>>> >>>> diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c >>>> index 65e7e56005b8..0d7f5f655fd8 100644 >>>> --- a/fs/ramfs/inode.c >>>> +++ b/fs/ramfs/inode.c >>>> @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context >>>> *fc, struct fs_parameter *param) >>>> struct ramfs_fs_info *fsi = fc->s_fs_info; >>>> int opt; >>>> + opt = vfs_parse_fs_param_source(fc, param); >>>> + if (opt != -ENOPARAM) >>>> + return opt; >>>> + >>>> opt = fs_parse(fc, ramfs_fs_parameters, param, &result); >>>> if (opt < 0) { >>>> /* >>>> >> >> (top-posting repaired) >> >>> Hi, this patch seems still leave in linux-next, should we pull it to >>> mainline? >> >> I was hoping for a comment from Al? > > Hi, Al, > > Can you help to review this patch... Hi, Al, Sorry for the noise again, can you help to give some comments for this patch. > > Thanks, > Kun. > >> . >> > .
Ping... 在 2021/9/13 9:10, yangerkun 写道: > > > 在 2021/9/9 16:37, yangerkun 写道: >> >> >> 在 2021/9/9 6:39, Andrew Morton 写道: >>> On Wed, 8 Sep 2021 16:56:25 +0800 yangerkun <yangerkun@huawei.com> >>> wrote: >>> >>>> 在 2021/8/11 20:28, yangerkun 写道: >>>>> ramfs_parse_param does not parse key "source", and will convert >>>>> -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in >>>>> vfs_parse_fs_param, which lead always "none" mount source for >>>>> ramfs. Fix >>>>> it by parse "source" in ramfs_parse_param. >>>>> >>>>> Signed-off-by: yangerkun <yangerkun@huawei.com> >>>>> --- >>>>> fs/ramfs/inode.c | 4 ++++ >>>>> 1 file changed, 4 insertions(+) >>>>> >>>>> diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c >>>>> index 65e7e56005b8..0d7f5f655fd8 100644 >>>>> --- a/fs/ramfs/inode.c >>>>> +++ b/fs/ramfs/inode.c >>>>> @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context >>>>> *fc, struct fs_parameter *param) >>>>> struct ramfs_fs_info *fsi = fc->s_fs_info; >>>>> int opt; >>>>> + opt = vfs_parse_fs_param_source(fc, param); >>>>> + if (opt != -ENOPARAM) >>>>> + return opt; >>>>> + >>>>> opt = fs_parse(fc, ramfs_fs_parameters, param, &result); >>>>> if (opt < 0) { >>>>> /* >>>>> >>> >>> (top-posting repaired) >>> >>>> Hi, this patch seems still leave in linux-next, should we pull it to >>>> mainline? >>> >>> I was hoping for a comment from Al? >> >> Hi, Al, >> >> Can you help to review this patch... > > Hi, Al, > > Sorry for the noise again, can you help to give some comments for this > patch. > >> >> Thanks, >> Kun. >> >>> . >>> >> . > .
Ping again... 在 2021/9/18 15:08, yangerkun 写道: > Ping... > > 在 2021/9/13 9:10, yangerkun 写道: >> >> >> 在 2021/9/9 16:37, yangerkun 写道: >>> >>> >>> 在 2021/9/9 6:39, Andrew Morton 写道: >>>> On Wed, 8 Sep 2021 16:56:25 +0800 yangerkun <yangerkun@huawei.com> >>>> wrote: >>>> >>>>> 在 2021/8/11 20:28, yangerkun 写道: >>>>>> ramfs_parse_param does not parse key "source", and will convert >>>>>> -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in >>>>>> vfs_parse_fs_param, which lead always "none" mount source for >>>>>> ramfs. Fix >>>>>> it by parse "source" in ramfs_parse_param. >>>>>> >>>>>> Signed-off-by: yangerkun <yangerkun@huawei.com> >>>>>> --- >>>>>> fs/ramfs/inode.c | 4 ++++ >>>>>> 1 file changed, 4 insertions(+) >>>>>> >>>>>> diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c >>>>>> index 65e7e56005b8..0d7f5f655fd8 100644 >>>>>> --- a/fs/ramfs/inode.c >>>>>> +++ b/fs/ramfs/inode.c >>>>>> @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct >>>>>> fs_context *fc, struct fs_parameter *param) >>>>>> struct ramfs_fs_info *fsi = fc->s_fs_info; >>>>>> int opt; >>>>>> + opt = vfs_parse_fs_param_source(fc, param); >>>>>> + if (opt != -ENOPARAM) >>>>>> + return opt; >>>>>> + >>>>>> opt = fs_parse(fc, ramfs_fs_parameters, param, &result); >>>>>> if (opt < 0) { >>>>>> /* >>>>>> >>>> >>>> (top-posting repaired) >>>> >>>>> Hi, this patch seems still leave in linux-next, should we pull it to >>>>> mainline? >>>> >>>> I was hoping for a comment from Al? >>> >>> Hi, Al, >>> >>> Can you help to review this patch... >> >> Hi, Al, >> >> Sorry for the noise again, can you help to give some comments for this >> patch. >> >>> >>> Thanks, >>> Kun. >>> >>>> . >>>> >>> . >> . > .
On Wed, Aug 11, 2021 at 08:28:11PM +0800, yangerkun wrote: > ramfs_parse_param does not parse key "source", and will convert > -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in > vfs_parse_fs_param, which lead always "none" mount source for ramfs. Fix > it by parse "source" in ramfs_parse_param. > > Signed-off-by: yangerkun <yangerkun@huawei.com> > --- > fs/ramfs/inode.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c > index 65e7e56005b8..0d7f5f655fd8 100644 > --- a/fs/ramfs/inode.c > +++ b/fs/ramfs/inode.c > @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context *fc, struct fs_parameter *param) > struct ramfs_fs_info *fsi = fc->s_fs_info; > int opt; > > + opt = vfs_parse_fs_param_source(fc, param); > + if (opt != -ENOPARAM) > + return opt; > + > opt = fs_parse(fc, ramfs_fs_parameters, param, &result); > if (opt < 0) { > /* Umm... If anything, I would rather call that thing *after* fs_parse() gives negative, similar to what kernel/cgroup/cgroup-v1.c does.
在 2021/9/24 12:35, Al Viro 写道: > On Wed, Aug 11, 2021 at 08:28:11PM +0800, yangerkun wrote: >> ramfs_parse_param does not parse key "source", and will convert >> -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in >> vfs_parse_fs_param, which lead always "none" mount source for ramfs. Fix >> it by parse "source" in ramfs_parse_param. >> >> Signed-off-by: yangerkun <yangerkun@huawei.com> >> --- >> fs/ramfs/inode.c | 4 ++++ >> 1 file changed, 4 insertions(+) >> >> diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c >> index 65e7e56005b8..0d7f5f655fd8 100644 >> --- a/fs/ramfs/inode.c >> +++ b/fs/ramfs/inode.c >> @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context *fc, struct fs_parameter *param) >> struct ramfs_fs_info *fsi = fc->s_fs_info; >> int opt; >> >> + opt = vfs_parse_fs_param_source(fc, param); >> + if (opt != -ENOPARAM) >> + return opt; >> + >> opt = fs_parse(fc, ramfs_fs_parameters, param, &result); >> if (opt < 0) { >> /* > > Umm... If anything, I would rather call that thing *after* > fs_parse() gives negative, similar to what kernel/cgroup/cgroup-v1.c > does. Thanks for your advise. Will do it next version. > . >
diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c index 65e7e56005b8..0d7f5f655fd8 100644 --- a/fs/ramfs/inode.c +++ b/fs/ramfs/inode.c @@ -202,6 +202,10 @@ static int ramfs_parse_param(struct fs_context *fc, struct fs_parameter *param) struct ramfs_fs_info *fsi = fc->s_fs_info; int opt; + opt = vfs_parse_fs_param_source(fc, param); + if (opt != -ENOPARAM) + return opt; + opt = fs_parse(fc, ramfs_fs_parameters, param, &result); if (opt < 0) { /*
ramfs_parse_param does not parse key "source", and will convert -ENOPARAM to 0. This will skip vfs_parse_fs_param_source in vfs_parse_fs_param, which lead always "none" mount source for ramfs. Fix it by parse "source" in ramfs_parse_param. Signed-off-by: yangerkun <yangerkun@huawei.com> --- fs/ramfs/inode.c | 4 ++++ 1 file changed, 4 insertions(+)