| Message ID | 20180530021311.25869-1-cgxu519@gmx.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
> On May 30, 2018, at 10:13, Chengguang Xu <cgxu519@gmx.com> wrote: > > Fix improper rasize alignment. > > Based on code, rasize(max readahead) actually aligns to PAGESIZE not > 1024 which is documented in below link, should fix the document as well. > > docs.ceph.com/docs/master/man/8/mount.ceph/ > > Signed-off-by: Chengguang Xu <cgxu519@gmx.com> > --- > fs/ceph/super.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/fs/ceph/super.c b/fs/ceph/super.c > index b33082e6878f..465861959a6b 100644 > --- a/fs/ceph/super.c > +++ b/fs/ceph/super.c > @@ -268,7 +268,7 @@ static int parse_fsopt_token(char *c, void *private) > case Opt_rasize: > if (intval < 0) > return -EINVAL; > - fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE); > + fsopt->rasize = ALIGN(intval, PAGE_SIZE); if intval is 2048, ALIGN(intval, PAGE_SIZE) is 0. why do you want to make this change? > break; > case Opt_caps_wanted_delay_min: > if (intval < 1) > -- > 2.17.0 > -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> 在 2018年5月30日，上午10:16，Yan, Zheng <zyan@redhat.com> 写道： > > > >> On May 30, 2018, at 10:13, Chengguang Xu <cgxu519@gmx.com> wrote: >> >> Fix improper rasize alignment. >> >> Based on code, rasize(max readahead) actually aligns to PAGESIZE not >> 1024 which is documented in below link, should fix the document as well. >> >> docs.ceph.com/docs/master/man/8/mount.ceph/ >> >> Signed-off-by: Chengguang Xu <cgxu519@gmx.com> >> --- >> fs/ceph/super.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/fs/ceph/super.c b/fs/ceph/super.c >> index b33082e6878f..465861959a6b 100644 >> --- a/fs/ceph/super.c >> +++ b/fs/ceph/super.c >> @@ -268,7 +268,7 @@ static int parse_fsopt_token(char *c, void *private) >> case Opt_rasize: >> if (intval < 0) >> return -EINVAL; >> - fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE); >> + fsopt->rasize = ALIGN(intval, PAGE_SIZE); > > if intval is 2048, ALIGN(intval, PAGE_SIZE) is 0. why do you want to make this change? Hi Yan, Thanks for you quick reply. I think it aligns up, so it will be 4096 unless you specify rasize=0. The motivation of this patch is avoiding confusion about specifying rasize. On currently logic: when I specify rasize=0~1 then it will be 4096. when I specify rasize=2~4097 then it will be 8192. I think the result will easily cause user’s confusion. Thanks, Chengguang. -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Then maybe it would be
fsopt->rasize = ALIGN_DOWN(intval + PAGE_SIZE - 1, PAGE_SIZE);
Thanx
On 05/30/2018 10:45 AM, cgxu519@gmx.com wrote:
> fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE);
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
> On May 30, 2018, at 10:45, cgxu519@gmx.com wrote: > >> >> 在 2018年5月30日，上午10:16，Yan, Zheng <zyan@redhat.com> 写道： >> >> >> >>> On May 30, 2018, at 10:13, Chengguang Xu <cgxu519@gmx.com> wrote: >>> >>> Fix improper rasize alignment. >>> >>> Based on code, rasize(max readahead) actually aligns to PAGESIZE not >>> 1024 which is documented in below link, should fix the document as well. >>> >>> docs.ceph.com/docs/master/man/8/mount.ceph/ >>> >>> Signed-off-by: Chengguang Xu <cgxu519@gmx.com> >>> --- >>> fs/ceph/super.c | 2 +- >>> 1 file changed, 1 insertion(+), 1 deletion(-) >>> >>> diff --git a/fs/ceph/super.c b/fs/ceph/super.c >>> index b33082e6878f..465861959a6b 100644 >>> --- a/fs/ceph/super.c >>> +++ b/fs/ceph/super.c >>> @@ -268,7 +268,7 @@ static int parse_fsopt_token(char *c, void *private) >>> case Opt_rasize: >>> if (intval < 0) >>> return -EINVAL; >>> - fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE); >>> + fsopt->rasize = ALIGN(intval, PAGE_SIZE); >> >> if intval is 2048, ALIGN(intval, PAGE_SIZE) is 0. why do you want to make this change? > > Hi Yan, > > Thanks for you quick reply. > > I think it aligns up, so it will be 4096 unless you specify rasize=0. > Ok, thanks for explanation. I will add it to our testing branch Yan, Zheng > The motivation of this patch is avoiding confusion about specifying rasize. > > On currently logic: > when I specify rasize=0~1 then it will be 4096. > when I specify rasize=2~4097 then it will be 8192. > > I think the result will easily cause user’s confusion. > > Thanks, > Chengguang. -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
It’s trivial but I don’t see any benefit from it，do you? Considering rsize & wsize, I would rather to keep same form. > 在 2018年5月30日，上午11:18，Dongsheng Yang <dongsheng.yang@easystack.cn> 写道： > > > Then maybe it would be > > fsopt->rasize = ALIGN_DOWN(intval + PAGE_SIZE - 1, PAGE_SIZE); > > Thanx > > > On 05/30/2018 10:45 AM, cgxu519@gmx.com wrote: >> fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE); > > -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Ha, that's okey to use ALIGN On 05/30/2018 11:26 AM, cgxu519@gmx.com wrote: > It’s trivial but I don’t see any benefit from it，do you? > > Considering rsize & wsize, I would rather to keep same form. > > > >> 在 2018年5月30日，上午11:18，Dongsheng Yang <dongsheng.yang@easystack.cn> 写道： >> >> >> Then maybe it would be >> >> fsopt->rasize = ALIGN_DOWN(intval + PAGE_SIZE - 1, PAGE_SIZE); >> >> Thanx >> >> >> On 05/30/2018 10:45 AM, cgxu519@gmx.com wrote: >>> fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE); >> > -- > To unsubscribe from this list: send the line "unsubscribe ceph-devel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/fs/ceph/super.c b/fs/ceph/super.c index b33082e6878f..465861959a6b 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -268,7 +268,7 @@ static int parse_fsopt_token(char *c, void *private) case Opt_rasize: if (intval < 0) return -EINVAL; - fsopt->rasize = ALIGN(intval + PAGE_SIZE - 1, PAGE_SIZE); + fsopt->rasize = ALIGN(intval, PAGE_SIZE); break; case Opt_caps_wanted_delay_min: if (intval < 1)
Fix improper rasize alignment. Based on code, rasize(max readahead) actually aligns to PAGESIZE not 1024 which is documented in below link, should fix the document as well. docs.ceph.com/docs/master/man/8/mount.ceph/ Signed-off-by: Chengguang Xu <cgxu519@gmx.com> --- fs/ceph/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)