* [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim @ 2017-10-11 13:51 Yunlong Song 2017-10-12 23:23 ` Jaegeuk Kim ` (2 more replies) 0 siblings, 3 replies; 27+ messages in thread From: Yunlong Song @ 2017-10-11 13:51 UTC (permalink / raw) To: jaegeuk, chao, yuchao0, yunlong.song, yunlong.song Cc: miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel This can help us to debug on some corner case. Signed-off-by: Yunlong Song <yunlong.song@huawei.com> --- fs/f2fs/gc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 197ebf4..960503e 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, .ilist = LIST_HEAD_INIT(gc_list.ilist), .iroot = RADIX_TREE_INIT(GFP_NOFS), }; + bool need_gc = false; trace_f2fs_gc_begin(sbi->sb, sync, background, get_pages(sbi, F2FS_DIRTY_NODES), @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, if (ret) goto stop; } - if (has_not_enough_free_secs(sbi, 0, 0)) + if (has_not_enough_free_secs(sbi, 0, 0)) { gc_type = FG_GC; + need_gc = true; + } } /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, goto stop; } if (!__get_victim(sbi, &segno, gc_type)) { + f2fs_bug_on(sbi, !total_freed && need_gc); ret = -ENODATA; goto stop; } -- 1.8.5.2 ^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-11 13:51 [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim Yunlong Song @ 2017-10-12 23:23 ` Jaegeuk Kim 2017-10-13 1:09 ` Yunlong Song 2017-10-13 11:09 ` [f2fs-dev] " Chao Yu 2017-10-13 13:31 ` [PATCH v2] " Yunlong Song 2 siblings, 1 reply; 27+ messages in thread From: Jaegeuk Kim @ 2017-10-12 23:23 UTC (permalink / raw) To: Yunlong Song Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel Hi Yunlong, On 10/11, Yunlong Song wrote: > This can help us to debug on some corner case. > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > --- > fs/f2fs/gc.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > index 197ebf4..960503e 100644 > --- a/fs/f2fs/gc.c > +++ b/fs/f2fs/gc.c > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > .ilist = LIST_HEAD_INIT(gc_list.ilist), > .iroot = RADIX_TREE_INIT(GFP_NOFS), > }; > + bool need_gc = false; > > trace_f2fs_gc_begin(sbi->sb, sync, background, > get_pages(sbi, F2FS_DIRTY_NODES), > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > if (ret) > goto stop; > } > - if (has_not_enough_free_secs(sbi, 0, 0)) > + if (has_not_enough_free_secs(sbi, 0, 0)) { > gc_type = FG_GC; > + need_gc = true; > + } > } > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > goto stop; > } > if (!__get_victim(sbi, &segno, gc_type)) { > + f2fs_bug_on(sbi, !total_freed && need_gc); How about this, if this is for debugging purpose? if (!total_freed && need_gc) f2fs_msg(...); It's not a good way to define a value for f2fs_bug_on(), since it is only activated by CHECK_FS config. Thanks, > ret = -ENODATA; > goto stop; > } > -- > 1.8.5.2 ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-12 23:23 ` Jaegeuk Kim @ 2017-10-13 1:09 ` Yunlong Song 2017-10-13 2:29 ` Jaegeuk Kim 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-10-13 1:09 UTC (permalink / raw) To: Jaegeuk Kim Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel Hi, Jay, I think it should not happen when need_gc == true but total_freed ==0, so I add it as bug_on to let it panic at once. And even CHECK_FS is closed, this can also printk WARNING message for notice. On 2017/10/13 7:23, Jaegeuk Kim wrote: > Hi Yunlong, > > On 10/11, Yunlong Song wrote: >> This can help us to debug on some corner case. >> >> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >> --- >> fs/f2fs/gc.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >> index 197ebf4..960503e 100644 >> --- a/fs/f2fs/gc.c >> +++ b/fs/f2fs/gc.c >> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> .ilist = LIST_HEAD_INIT(gc_list.ilist), >> .iroot = RADIX_TREE_INIT(GFP_NOFS), >> }; >> + bool need_gc = false; >> >> trace_f2fs_gc_begin(sbi->sb, sync, background, >> get_pages(sbi, F2FS_DIRTY_NODES), >> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> if (ret) >> goto stop; >> } >> - if (has_not_enough_free_secs(sbi, 0, 0)) >> + if (has_not_enough_free_secs(sbi, 0, 0)) { >> gc_type = FG_GC; >> + need_gc = true; >> + } >> } >> >> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> goto stop; >> } >> if (!__get_victim(sbi, &segno, gc_type)) { >> + f2fs_bug_on(sbi, !total_freed && need_gc); > How about this, if this is for debugging purpose? > > if (!total_freed && need_gc) > f2fs_msg(...); > > It's not a good way to define a value for f2fs_bug_on(), since it is only > activated by CHECK_FS config. > > Thanks, > >> ret = -ENODATA; >> goto stop; >> } >> -- >> 1.8.5.2 > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-13 1:09 ` Yunlong Song @ 2017-10-13 2:29 ` Jaegeuk Kim 0 siblings, 0 replies; 27+ messages in thread From: Jaegeuk Kim @ 2017-10-13 2:29 UTC (permalink / raw) To: Yunlong Song Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 10/13, Yunlong Song wrote: > Hi, Jay, > I think it should not happen when need_gc == true but total_freed ==0, > so I add it as bug_on to let it panic at once. And even CHECK_FS is closed, > this can also printk WARNING message for notice. Ah, got it. Merged. :) Thanks, > > On 2017/10/13 7:23, Jaegeuk Kim wrote: > > Hi Yunlong, > > > > On 10/11, Yunlong Song wrote: > > > This can help us to debug on some corner case. > > > > > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > > > --- > > > fs/f2fs/gc.c | 6 +++++- > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > > > index 197ebf4..960503e 100644 > > > --- a/fs/f2fs/gc.c > > > +++ b/fs/f2fs/gc.c > > > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > .ilist = LIST_HEAD_INIT(gc_list.ilist), > > > .iroot = RADIX_TREE_INIT(GFP_NOFS), > > > }; > > > + bool need_gc = false; > > > trace_f2fs_gc_begin(sbi->sb, sync, background, > > > get_pages(sbi, F2FS_DIRTY_NODES), > > > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > if (ret) > > > goto stop; > > > } > > > - if (has_not_enough_free_secs(sbi, 0, 0)) > > > + if (has_not_enough_free_secs(sbi, 0, 0)) { > > > gc_type = FG_GC; > > > + need_gc = true; > > > + } > > > } > > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > > > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > goto stop; > > > } > > > if (!__get_victim(sbi, &segno, gc_type)) { > > > + f2fs_bug_on(sbi, !total_freed && need_gc); > > How about this, if this is for debugging purpose? > > > > if (!total_freed && need_gc) > > f2fs_msg(...); > > > > It's not a good way to define a value for f2fs_bug_on(), since it is only > > activated by CHECK_FS config. > > > > Thanks, > > > > > ret = -ENODATA; > > > goto stop; > > > } > > > -- > > > 1.8.5.2 > > . > > > > -- > Thanks, > Yunlong Song > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-11 13:51 [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim Yunlong Song 2017-10-12 23:23 ` Jaegeuk Kim @ 2017-10-13 11:09 ` Chao Yu 2017-10-13 11:20 ` Yunlong Song 2017-10-13 13:31 ` [PATCH v2] " Yunlong Song 2 siblings, 1 reply; 27+ messages in thread From: Chao Yu @ 2017-10-13 11:09 UTC (permalink / raw) To: Yunlong Song, jaegeuk, yuchao0, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel On 2017/10/11 21:51, Yunlong Song wrote: > This can help us to debug on some corner case. > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > --- > fs/f2fs/gc.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > index 197ebf4..960503e 100644 > --- a/fs/f2fs/gc.c > +++ b/fs/f2fs/gc.c > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > .ilist = LIST_HEAD_INIT(gc_list.ilist), > .iroot = RADIX_TREE_INIT(GFP_NOFS), > }; > + bool need_gc = false; How about changing variable name to need_fggc for better readability? Thanks, > > trace_f2fs_gc_begin(sbi->sb, sync, background, > get_pages(sbi, F2FS_DIRTY_NODES), > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > if (ret) > goto stop; > } > - if (has_not_enough_free_secs(sbi, 0, 0)) > + if (has_not_enough_free_secs(sbi, 0, 0)) { > gc_type = FG_GC; > + need_gc = true; > + } > } > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > goto stop; > } > if (!__get_victim(sbi, &segno, gc_type)) { > + f2fs_bug_on(sbi, !total_freed && need_gc); > ret = -ENODATA; > goto stop; > } > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-13 11:09 ` [f2fs-dev] " Chao Yu @ 2017-10-13 11:20 ` Yunlong Song 2017-10-13 12:24 ` Chao Yu 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-10-13 11:20 UTC (permalink / raw) To: Chao Yu, jaegeuk, yuchao0, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel Yep, both are OK, since I do not care it is foreground or background. Here it just needs to do gc. On 2017/10/13 19:09, Chao Yu wrote: > On 2017/10/11 21:51, Yunlong Song wrote: >> This can help us to debug on some corner case. >> >> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >> --- >> fs/f2fs/gc.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >> index 197ebf4..960503e 100644 >> --- a/fs/f2fs/gc.c >> +++ b/fs/f2fs/gc.c >> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> .ilist = LIST_HEAD_INIT(gc_list.ilist), >> .iroot = RADIX_TREE_INIT(GFP_NOFS), >> }; >> + bool need_gc = false; > How about changing variable name to need_fggc for better readability? > > Thanks, > >> >> trace_f2fs_gc_begin(sbi->sb, sync, background, >> get_pages(sbi, F2FS_DIRTY_NODES), >> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> if (ret) >> goto stop; >> } >> - if (has_not_enough_free_secs(sbi, 0, 0)) >> + if (has_not_enough_free_secs(sbi, 0, 0)) { >> gc_type = FG_GC; >> + need_gc = true; >> + } >> } >> >> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> goto stop; >> } >> if (!__get_victim(sbi, &segno, gc_type)) { >> + f2fs_bug_on(sbi, !total_freed && need_gc); >> ret = -ENODATA; >> goto stop; >> } >> > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-13 11:20 ` Yunlong Song @ 2017-10-13 12:24 ` Chao Yu 0 siblings, 0 replies; 27+ messages in thread From: Chao Yu @ 2017-10-13 12:24 UTC (permalink / raw) To: Yunlong Song, jaegeuk, yuchao0, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel On 2017/10/13 19:20, Yunlong Song wrote: > Yep, both are OK, since I do not care it is foreground or background. Here it just needs to do gc. But you only set need_gc as true for foreground GC case. Thanks, > > On 2017/10/13 19:09, Chao Yu wrote: >> On 2017/10/11 21:51, Yunlong Song wrote: >>> This can help us to debug on some corner case. >>> >>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>> --- >>> fs/f2fs/gc.c | 6 +++++- >>> 1 file changed, 5 insertions(+), 1 deletion(-) >>> >>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>> index 197ebf4..960503e 100644 >>> --- a/fs/f2fs/gc.c >>> +++ b/fs/f2fs/gc.c >>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>> }; >>> + bool need_gc = false; >> How about changing variable name to need_fggc for better readability? >> >> Thanks, >> >>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>> get_pages(sbi, F2FS_DIRTY_NODES), >>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>> if (ret) >>> goto stop; >>> } >>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>> gc_type = FG_GC; >>> + need_gc = true; >>> + } >>> } >>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>> goto stop; >>> } >>> if (!__get_victim(sbi, &segno, gc_type)) { >>> + f2fs_bug_on(sbi, !total_freed && need_gc); >>> ret = -ENODATA; >>> goto stop; >>> } >>> >> . >> > ^ permalink raw reply [flat|nested] 27+ messages in thread
* [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-11 13:51 [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim Yunlong Song 2017-10-12 23:23 ` Jaegeuk Kim 2017-10-13 11:09 ` [f2fs-dev] " Chao Yu @ 2017-10-13 13:31 ` Yunlong Song 2017-10-14 0:17 ` [f2fs-dev] " Chao Yu ` (2 more replies) 2 siblings, 3 replies; 27+ messages in thread From: Yunlong Song @ 2017-10-13 13:31 UTC (permalink / raw) To: jaegeuk, chao, yuchao0, yunlong.song, yunlong.song Cc: miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel This can help us to debug on some corner case. Signed-off-by: Yunlong Song <yunlong.song@huawei.com> Signed-off-by: Chao Yu <yuchao0@huawei.com> --- fs/f2fs/gc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 197ebf4..2b03202 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, .ilist = LIST_HEAD_INIT(gc_list.ilist), .iroot = RADIX_TREE_INIT(GFP_NOFS), }; + bool need_fggc = false; trace_f2fs_gc_begin(sbi->sb, sync, background, get_pages(sbi, F2FS_DIRTY_NODES), @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, if (ret) goto stop; } - if (has_not_enough_free_secs(sbi, 0, 0)) + if (has_not_enough_free_secs(sbi, 0, 0)) { gc_type = FG_GC; + need_fggc = true; + } } /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, goto stop; } if (!__get_victim(sbi, &segno, gc_type)) { + f2fs_bug_on(sbi, !total_freed && need_fggc); ret = -ENODATA; goto stop; } -- 1.8.5.2 ^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-13 13:31 ` [PATCH v2] " Yunlong Song @ 2017-10-14 0:17 ` Chao Yu 2017-10-14 12:34 ` Yunlong Song 2017-11-03 3:28 ` Yunlong Song 2017-11-03 3:44 ` Jaegeuk Kim 2 siblings, 1 reply; 27+ messages in thread From: Chao Yu @ 2017-10-14 0:17 UTC (permalink / raw) To: Yunlong Song, jaegeuk, yuchao0, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel On 2017/10/13 21:31, Yunlong Song wrote: > This can help us to debug on some corner case. I can hit this bugon with generic/015 of fstest easily, could have a look at this? Thanks, > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > Signed-off-by: Chao Yu <yuchao0@huawei.com> > --- > fs/f2fs/gc.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > index 197ebf4..2b03202 100644 > --- a/fs/f2fs/gc.c > +++ b/fs/f2fs/gc.c > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > .ilist = LIST_HEAD_INIT(gc_list.ilist), > .iroot = RADIX_TREE_INIT(GFP_NOFS), > }; > + bool need_fggc = false; > > trace_f2fs_gc_begin(sbi->sb, sync, background, > get_pages(sbi, F2FS_DIRTY_NODES), > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > if (ret) > goto stop; > } > - if (has_not_enough_free_secs(sbi, 0, 0)) > + if (has_not_enough_free_secs(sbi, 0, 0)) { > gc_type = FG_GC; > + need_fggc = true; > + } > } > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > goto stop; > } > if (!__get_victim(sbi, &segno, gc_type)) { > + f2fs_bug_on(sbi, !total_freed && need_fggc); > ret = -ENODATA; > goto stop; > } > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-14 0:17 ` [f2fs-dev] " Chao Yu @ 2017-10-14 12:34 ` Yunlong Song 2017-10-16 3:25 ` Chao Yu 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-10-14 12:34 UTC (permalink / raw) To: Chao Yu, jaegeuk, yuchao0, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel Do you mean check out-of-space test? I have tried that but no bugon. On 2017/10/14 8:17, Chao Yu wrote: > On 2017/10/13 21:31, Yunlong Song wrote: >> This can help us to debug on some corner case. > I can hit this bugon with generic/015 of fstest easily, could have a look at > this? > > Thanks, > >> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >> Signed-off-by: Chao Yu <yuchao0@huawei.com> >> --- >> fs/f2fs/gc.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >> index 197ebf4..2b03202 100644 >> --- a/fs/f2fs/gc.c >> +++ b/fs/f2fs/gc.c >> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> .ilist = LIST_HEAD_INIT(gc_list.ilist), >> .iroot = RADIX_TREE_INIT(GFP_NOFS), >> }; >> + bool need_fggc = false; >> >> trace_f2fs_gc_begin(sbi->sb, sync, background, >> get_pages(sbi, F2FS_DIRTY_NODES), >> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> if (ret) >> goto stop; >> } >> - if (has_not_enough_free_secs(sbi, 0, 0)) >> + if (has_not_enough_free_secs(sbi, 0, 0)) { >> gc_type = FG_GC; >> + need_fggc = true; >> + } >> } >> >> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> goto stop; >> } >> if (!__get_victim(sbi, &segno, gc_type)) { >> + f2fs_bug_on(sbi, !total_freed && need_fggc); >> ret = -ENODATA; >> goto stop; >> } >> > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-14 12:34 ` Yunlong Song @ 2017-10-16 3:25 ` Chao Yu 2017-10-31 1:32 ` Yunlong Song 0 siblings, 1 reply; 27+ messages in thread From: Chao Yu @ 2017-10-16 3:25 UTC (permalink / raw) To: Yunlong Song, Chao Yu, jaegeuk, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel On 2017/10/14 20:34, Yunlong Song wrote: > Do you mean check out-of-space test? I have tried that but no bugon. Yes, test recent f2fs codes with kernel 4.13.0-rc1+ in VM, FYI: kernel BUG at gc.c:1034! invalid opcode: 0000 [#1] SMP Hardware name: Xen HVM domU, BIOS 4.1.2_115-900.260_ 11/06/2015 RIP: 0010:f2fs_gc+0x6e5/0x6f0 [f2fs] RSP: 0018:ffffc90004af7b40 EFLAGS: 00010202 RAX: ffff8801b0a15940 RBX: 0000000000000000 RCX: 0000000000000000 RDX: ffff8801b0a15940 RSI: ffff8801978d5f00 RDI: ffff880128148048 RBP: ffffc90004af7c38 R08: ffff8801978d5f00 R09: 0000000000000003 R10: 0000000000000003 R11: ffff8800060703a0 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000001 R15: ffff8801b4279800 FS: 00007f23493cb740(0000) GS:ffff880216f00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007ffd05402ff8 CR3: 00000001bffb3000 CR4: 00000000001406e0 Call Trace: f2fs_balance_fs+0x123/0x140 [f2fs] f2fs_create+0x130/0x240 [f2fs] path_openat+0xee7/0x1360 do_filp_open+0x7e/0xd0 do_sys_open+0x115/0x1f0 SyS_open+0x1e/0x20 do_syscall_64+0x6e/0x160 entry_SYSCALL64_slow_path+0x25/0x25 Thanks, > > On 2017/10/14 8:17, Chao Yu wrote: >> On 2017/10/13 21:31, Yunlong Song wrote: >>> This can help us to debug on some corner case. >> I can hit this bugon with generic/015 of fstest easily, could have a look at >> this? >> >> Thanks, >> >>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>> --- >>> fs/f2fs/gc.c | 6 +++++- >>> 1 file changed, 5 insertions(+), 1 deletion(-) >>> >>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>> index 197ebf4..2b03202 100644 >>> --- a/fs/f2fs/gc.c >>> +++ b/fs/f2fs/gc.c >>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>> }; >>> + bool need_fggc = false; >>> >>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>> get_pages(sbi, F2FS_DIRTY_NODES), >>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>> if (ret) >>> goto stop; >>> } >>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>> gc_type = FG_GC; >>> + need_fggc = true; >>> + } >>> } >>> >>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>> goto stop; >>> } >>> if (!__get_victim(sbi, &segno, gc_type)) { >>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>> ret = -ENODATA; >>> goto stop; >>> } >>> >> . >> > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-16 3:25 ` Chao Yu @ 2017-10-31 1:32 ` Yunlong Song 2017-10-31 3:17 ` Chao Yu 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-10-31 1:32 UTC (permalink / raw) To: Chao Yu, Chao Yu, jaegeuk, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel I think there may be bugs somewhere, since no victim is selected but it really needs gc. What is the size of the data image? On 2017/10/16 11:25, Chao Yu wrote: > On 2017/10/14 20:34, Yunlong Song wrote: >> Do you mean check out-of-space test? I have tried that but no bugon. > Yes, test recent f2fs codes with kernel 4.13.0-rc1+ in VM, FYI: > > kernel BUG at gc.c:1034! > invalid opcode: 0000 [#1] SMP > Hardware name: Xen HVM domU, BIOS 4.1.2_115-900.260_ 11/06/2015 > RIP: 0010:f2fs_gc+0x6e5/0x6f0 [f2fs] > RSP: 0018:ffffc90004af7b40 EFLAGS: 00010202 > RAX: ffff8801b0a15940 RBX: 0000000000000000 RCX: 0000000000000000 > RDX: ffff8801b0a15940 RSI: ffff8801978d5f00 RDI: ffff880128148048 > RBP: ffffc90004af7c38 R08: ffff8801978d5f00 R09: 0000000000000003 > R10: 0000000000000003 R11: ffff8800060703a0 R12: 0000000000000000 > R13: 0000000000000000 R14: 0000000000000001 R15: ffff8801b4279800 > FS: 00007f23493cb740(0000) GS:ffff880216f00000(0000) knlGS:0000000000000000 > CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > CR2: 00007ffd05402ff8 CR3: 00000001bffb3000 CR4: 00000000001406e0 > Call Trace: > f2fs_balance_fs+0x123/0x140 [f2fs] > f2fs_create+0x130/0x240 [f2fs] > path_openat+0xee7/0x1360 > do_filp_open+0x7e/0xd0 > do_sys_open+0x115/0x1f0 > SyS_open+0x1e/0x20 > do_syscall_64+0x6e/0x160 > entry_SYSCALL64_slow_path+0x25/0x25 > > Thanks, > >> On 2017/10/14 8:17, Chao Yu wrote: >>> On 2017/10/13 21:31, Yunlong Song wrote: >>>> This can help us to debug on some corner case. >>> I can hit this bugon with generic/015 of fstest easily, could have a look at >>> this? >>> >>> Thanks, >>> >>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>> --- >>>> fs/f2fs/gc.c | 6 +++++- >>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>> >>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>> index 197ebf4..2b03202 100644 >>>> --- a/fs/f2fs/gc.c >>>> +++ b/fs/f2fs/gc.c >>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>> }; >>>> + bool need_fggc = false; >>>> >>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>> if (ret) >>>> goto stop; >>>> } >>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>> gc_type = FG_GC; >>>> + need_fggc = true; >>>> + } >>>> } >>>> >>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>> goto stop; >>>> } >>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>> ret = -ENODATA; >>>> goto stop; >>>> } >>>> >>> . >>> > > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-31 1:32 ` Yunlong Song @ 2017-10-31 3:17 ` Chao Yu 2017-10-31 3:30 ` Yunlong Song 0 siblings, 1 reply; 27+ messages in thread From: Chao Yu @ 2017-10-31 3:17 UTC (permalink / raw) To: Yunlong Song, Chao Yu, jaegeuk, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel On 2017/10/31 9:32, Yunlong Song wrote: > I think there may be bugs somewhere, since no victim is selected but it > really needs gc. > What is the size of the data image? I have providered the testcase, could you check that? >>>> I can hit this bugon with generic/015 of fstest easily, could have a look at >>>> this? Anyway, I have looked into this issue at last weekend, forgot to post related info. It shows that if the image size is very small, e.g. 50MB, current segment will encroach overprov and reserved segments, so free segment count will always be small than reserved segment threshold, result in triggering bugon in FGGC. [SB: 1] [CP: 2] [SIT: 2] [NAT: 2] [SSA: 1] [MAIN: 17(OverProv:14 Resv:12)] - Valid: 6 - Dirty: 0 - Prefree: 0 - Free: 11 (11) I modify mkfs.f2fs forcedly as below, which could let generic/015 passed: diff --git a/mkfs/f2fs_format.c b/mkfs/f2fs_format.c index 8a821d3a1747..774bbef73d9a 100644 --- a/mkfs/f2fs_format.c +++ b/mkfs/f2fs_format.c @@ -371,6 +371,12 @@ static int f2fs_prepare_super_block(void) (2 * (100 / c.overprovision + 1) + 6) * c.segs_per_sec; + if (get_sb(segment_count_main) * (100 - c.overprovision) / 100 - + c.reserved_segments < 6) { + c.overprovision = 50; + c.reserved_segments = get_sb(segment_count_main) - 6 * 2; + } + uuid_generate(sb->uuid); /* precompute checksum seed for metadata */ Thanks, > > On 2017/10/16 11:25, Chao Yu wrote: >> On 2017/10/14 20:34, Yunlong Song wrote: >>> Do you mean check out-of-space test? I have tried that but no bugon. >> Yes, test recent f2fs codes with kernel 4.13.0-rc1+ in VM, FYI: >> >> kernel BUG at gc.c:1034! >> invalid opcode: 0000 [#1] SMP >> Hardware name: Xen HVM domU, BIOS 4.1.2_115-900.260_ 11/06/2015 >> RIP: 0010:f2fs_gc+0x6e5/0x6f0 [f2fs] >> RSP: 0018:ffffc90004af7b40 EFLAGS: 00010202 >> RAX: ffff8801b0a15940 RBX: 0000000000000000 RCX: 0000000000000000 >> RDX: ffff8801b0a15940 RSI: ffff8801978d5f00 RDI: ffff880128148048 >> RBP: ffffc90004af7c38 R08: ffff8801978d5f00 R09: 0000000000000003 >> R10: 0000000000000003 R11: ffff8800060703a0 R12: 0000000000000000 >> R13: 0000000000000000 R14: 0000000000000001 R15: ffff8801b4279800 >> FS: 00007f23493cb740(0000) GS:ffff880216f00000(0000) knlGS:0000000000000000 >> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 >> CR2: 00007ffd05402ff8 CR3: 00000001bffb3000 CR4: 00000000001406e0 >> Call Trace: >> f2fs_balance_fs+0x123/0x140 [f2fs] >> f2fs_create+0x130/0x240 [f2fs] >> path_openat+0xee7/0x1360 >> do_filp_open+0x7e/0xd0 >> do_sys_open+0x115/0x1f0 >> SyS_open+0x1e/0x20 >> do_syscall_64+0x6e/0x160 >> entry_SYSCALL64_slow_path+0x25/0x25 >> >> Thanks, >> >>> On 2017/10/14 8:17, Chao Yu wrote: >>>> On 2017/10/13 21:31, Yunlong Song wrote: >>>>> This can help us to debug on some corner case. >>>> I can hit this bugon with generic/015 of fstest easily, could have a look at >>>> this? >>>> >>>> Thanks, >>>> >>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>> --- >>>>> fs/f2fs/gc.c | 6 +++++- >>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>> >>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>> index 197ebf4..2b03202 100644 >>>>> --- a/fs/f2fs/gc.c >>>>> +++ b/fs/f2fs/gc.c >>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>> }; >>>>> + bool need_fggc = false; >>>>> >>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>> if (ret) >>>>> goto stop; >>>>> } >>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>> gc_type = FG_GC; >>>>> + need_fggc = true; >>>>> + } >>>>> } >>>>> >>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>> goto stop; >>>>> } >>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>>> ret = -ENODATA; >>>>> goto stop; >>>>> } >>>>> >>>> . >>>> >> >> . >> > ^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [f2fs-dev] [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-31 3:17 ` Chao Yu @ 2017-10-31 3:30 ` Yunlong Song 0 siblings, 0 replies; 27+ messages in thread From: Yunlong Song @ 2017-10-31 3:30 UTC (permalink / raw) To: Chao Yu, Chao Yu, jaegeuk, yunlong.song Cc: linux-fsdevel, miaoxie, linux-kernel, linux-f2fs-devel So it seems it is really useful to add this bug_on in gc. On 2017/10/31 11:17, Chao Yu wrote: > On 2017/10/31 9:32, Yunlong Song wrote: >> I think there may be bugs somewhere, since no victim is selected but it >> really needs gc. >> What is the size of the data image? > I have providered the testcase, could you check that? > >>>>> I can hit this bugon with generic/015 of fstest easily, could have a look at >>>>> this? > Anyway, I have looked into this issue at last weekend, forgot to post related > info. It shows that if the image size is very small, e.g. 50MB, current > segment will encroach overprov and reserved segments, so free segment count will > always be small than reserved segment threshold, result in triggering bugon in > FGGC. > > [SB: 1] [CP: 2] [SIT: 2] [NAT: 2] [SSA: 1] [MAIN: 17(OverProv:14 Resv:12)] > > - Valid: 6 > - Dirty: 0 > - Prefree: 0 > - Free: 11 (11) > > I modify mkfs.f2fs forcedly as below, which could let generic/015 passed: > > diff --git a/mkfs/f2fs_format.c b/mkfs/f2fs_format.c > index 8a821d3a1747..774bbef73d9a 100644 > --- a/mkfs/f2fs_format.c > +++ b/mkfs/f2fs_format.c > @@ -371,6 +371,12 @@ static int f2fs_prepare_super_block(void) > (2 * (100 / c.overprovision + 1) + 6) > * c.segs_per_sec; > > + if (get_sb(segment_count_main) * (100 - c.overprovision) / 100 - > + c.reserved_segments < 6) { > + c.overprovision = 50; > + c.reserved_segments = get_sb(segment_count_main) - 6 * 2; > + } > + > uuid_generate(sb->uuid); > > /* precompute checksum seed for metadata */ > > Thanks, > >> On 2017/10/16 11:25, Chao Yu wrote: >>> On 2017/10/14 20:34, Yunlong Song wrote: >>>> Do you mean check out-of-space test? I have tried that but no bugon. >>> Yes, test recent f2fs codes with kernel 4.13.0-rc1+ in VM, FYI: >>> >>> kernel BUG at gc.c:1034! >>> invalid opcode: 0000 [#1] SMP >>> Hardware name: Xen HVM domU, BIOS 4.1.2_115-900.260_ 11/06/2015 >>> RIP: 0010:f2fs_gc+0x6e5/0x6f0 [f2fs] >>> RSP: 0018:ffffc90004af7b40 EFLAGS: 00010202 >>> RAX: ffff8801b0a15940 RBX: 0000000000000000 RCX: 0000000000000000 >>> RDX: ffff8801b0a15940 RSI: ffff8801978d5f00 RDI: ffff880128148048 >>> RBP: ffffc90004af7c38 R08: ffff8801978d5f00 R09: 0000000000000003 >>> R10: 0000000000000003 R11: ffff8800060703a0 R12: 0000000000000000 >>> R13: 0000000000000000 R14: 0000000000000001 R15: ffff8801b4279800 >>> FS: 00007f23493cb740(0000) GS:ffff880216f00000(0000) knlGS:0000000000000000 >>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 >>> CR2: 00007ffd05402ff8 CR3: 00000001bffb3000 CR4: 00000000001406e0 >>> Call Trace: >>> f2fs_balance_fs+0x123/0x140 [f2fs] >>> f2fs_create+0x130/0x240 [f2fs] >>> path_openat+0xee7/0x1360 >>> do_filp_open+0x7e/0xd0 >>> do_sys_open+0x115/0x1f0 >>> SyS_open+0x1e/0x20 >>> do_syscall_64+0x6e/0x160 >>> entry_SYSCALL64_slow_path+0x25/0x25 >>> >>> Thanks, >>> >>>> On 2017/10/14 8:17, Chao Yu wrote: >>>>> On 2017/10/13 21:31, Yunlong Song wrote: >>>>>> This can help us to debug on some corner case. >>>>> I can hit this bugon with generic/015 of fstest easily, could have a look at >>>>> this? >>>>> >>>>> Thanks, >>>>> >>>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>>> --- >>>>>> fs/f2fs/gc.c | 6 +++++- >>>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>>> >>>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>>> index 197ebf4..2b03202 100644 >>>>>> --- a/fs/f2fs/gc.c >>>>>> +++ b/fs/f2fs/gc.c >>>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>>> }; >>>>>> + bool need_fggc = false; >>>>>> >>>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>> if (ret) >>>>>> goto stop; >>>>>> } >>>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>>> gc_type = FG_GC; >>>>>> + need_fggc = true; >>>>>> + } >>>>>> } >>>>>> >>>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>> goto stop; >>>>>> } >>>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>>>> ret = -ENODATA; >>>>>> goto stop; >>>>>> } >>>>>> >>>>> . >>>>> >>> . >>> > > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-13 13:31 ` [PATCH v2] " Yunlong Song 2017-10-14 0:17 ` [f2fs-dev] " Chao Yu @ 2017-11-03 3:28 ` Yunlong Song 2017-11-03 3:44 ` Jaegeuk Kim 2 siblings, 0 replies; 27+ messages in thread From: Yunlong Song @ 2017-11-03 3:28 UTC (permalink / raw) To: jaegeuk, chao, yuchao0, yunlong.song Cc: miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel ping... On 2017/10/13 21:31, Yunlong Song wrote: > This can help us to debug on some corner case. > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > Signed-off-by: Chao Yu <yuchao0@huawei.com> > --- > fs/f2fs/gc.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > index 197ebf4..2b03202 100644 > --- a/fs/f2fs/gc.c > +++ b/fs/f2fs/gc.c > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > .ilist = LIST_HEAD_INIT(gc_list.ilist), > .iroot = RADIX_TREE_INIT(GFP_NOFS), > }; > + bool need_fggc = false; > > trace_f2fs_gc_begin(sbi->sb, sync, background, > get_pages(sbi, F2FS_DIRTY_NODES), > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > if (ret) > goto stop; > } > - if (has_not_enough_free_secs(sbi, 0, 0)) > + if (has_not_enough_free_secs(sbi, 0, 0)) { > gc_type = FG_GC; > + need_fggc = true; > + } > } > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > goto stop; > } > if (!__get_victim(sbi, &segno, gc_type)) { > + f2fs_bug_on(sbi, !total_freed && need_fggc); > ret = -ENODATA; > goto stop; > } -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-10-13 13:31 ` [PATCH v2] " Yunlong Song 2017-10-14 0:17 ` [f2fs-dev] " Chao Yu 2017-11-03 3:28 ` Yunlong Song @ 2017-11-03 3:44 ` Jaegeuk Kim 2017-11-03 15:08 ` 宋云龙 2017-11-06 1:12 ` Yunlong Song 2 siblings, 2 replies; 27+ messages in thread From: Jaegeuk Kim @ 2017-11-03 3:44 UTC (permalink / raw) To: Yunlong Song Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 10/13, Yunlong Song wrote: > This can help us to debug on some corner case. > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > Signed-off-by: Chao Yu <yuchao0@huawei.com> > --- > fs/f2fs/gc.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > index 197ebf4..2b03202 100644 > --- a/fs/f2fs/gc.c > +++ b/fs/f2fs/gc.c > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > .ilist = LIST_HEAD_INIT(gc_list.ilist), > .iroot = RADIX_TREE_INIT(GFP_NOFS), > }; > + bool need_fggc = false; > > trace_f2fs_gc_begin(sbi->sb, sync, background, > get_pages(sbi, F2FS_DIRTY_NODES), > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > if (ret) > goto stop; > } > - if (has_not_enough_free_secs(sbi, 0, 0)) > + if (has_not_enough_free_secs(sbi, 0, 0)) { > gc_type = FG_GC; > + need_fggc = true; > + } > } > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > goto stop; > } > if (!__get_victim(sbi, &segno, gc_type)) { > + f2fs_bug_on(sbi, !total_freed && need_fggc); Just like this? f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); > ret = -ENODATA; > goto stop; > } > -- > 1.8.5.2 ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-03 3:44 ` Jaegeuk Kim @ 2017-11-03 15:08 ` 宋云龙 2017-11-06 1:12 ` Yunlong Song 1 sibling, 0 replies; 27+ messages in thread From: 宋云龙 @ 2017-11-03 15:08 UTC (permalink / raw) To: Jaegeuk Kim Cc: Yunlong Song, chao, yuchao0, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel OK to me. > > >> On 10/13, Yunlong Song wrote: >> This can help us to debug on some corner case. >> >> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >> Signed-off-by: Chao Yu <yuchao0@huawei.com> >> --- >> fs/f2fs/gc.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >> index 197ebf4..2b03202 100644 >> --- a/fs/f2fs/gc.c >> +++ b/fs/f2fs/gc.c >> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> .ilist = LIST_HEAD_INIT(gc_list.ilist), >> .iroot = RADIX_TREE_INIT(GFP_NOFS), >> }; >> + bool need_fggc = false; >> >> trace_f2fs_gc_begin(sbi->sb, sync, background, >> get_pages(sbi, F2FS_DIRTY_NODES), >> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> if (ret) >> goto stop; >> } >> - if (has_not_enough_free_secs(sbi, 0, 0)) >> + if (has_not_enough_free_secs(sbi, 0, 0)) { >> gc_type = FG_GC; >> + need_fggc = true; >> + } >> } >> >> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> goto stop; >> } >> if (!__get_victim(sbi, &segno, gc_type)) { >> + f2fs_bug_on(sbi, !total_freed && need_fggc); > > Just like this? > > f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); Agree > >> ret = -ENODATA; >> goto stop; >> } >> -- >> 1.8.5.2 ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-03 3:44 ` Jaegeuk Kim 2017-11-03 15:08 ` 宋云龙 @ 2017-11-06 1:12 ` Yunlong Song 2017-11-07 2:38 ` Jaegeuk Kim 1 sibling, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-11-06 1:12 UTC (permalink / raw) To: Jaegeuk Kim Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel Agree. On 2017/11/3 11:44, Jaegeuk Kim wrote: > On 10/13, Yunlong Song wrote: >> This can help us to debug on some corner case. >> >> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >> Signed-off-by: Chao Yu <yuchao0@huawei.com> >> --- >> fs/f2fs/gc.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) >> >> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >> index 197ebf4..2b03202 100644 >> --- a/fs/f2fs/gc.c >> +++ b/fs/f2fs/gc.c >> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> .ilist = LIST_HEAD_INIT(gc_list.ilist), >> .iroot = RADIX_TREE_INIT(GFP_NOFS), >> }; >> + bool need_fggc = false; >> >> trace_f2fs_gc_begin(sbi->sb, sync, background, >> get_pages(sbi, F2FS_DIRTY_NODES), >> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> if (ret) >> goto stop; >> } >> - if (has_not_enough_free_secs(sbi, 0, 0)) >> + if (has_not_enough_free_secs(sbi, 0, 0)) { >> gc_type = FG_GC; >> + need_fggc = true; >> + } >> } >> >> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >> goto stop; >> } >> if (!__get_victim(sbi, &segno, gc_type)) { >> + f2fs_bug_on(sbi, !total_freed && need_fggc); > Just like this? That's OK. > > f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); > >> ret = -ENODATA; >> goto stop; >> } >> -- >> 1.8.5.2 > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-06 1:12 ` Yunlong Song @ 2017-11-07 2:38 ` Jaegeuk Kim 2017-11-07 2:40 ` Jaegeuk Kim 0 siblings, 1 reply; 27+ messages in thread From: Jaegeuk Kim @ 2017-11-07 2:38 UTC (permalink / raw) To: Yunlong Song Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 11/06, Yunlong Song wrote: > Agree. > > On 2017/11/3 11:44, Jaegeuk Kim wrote: > > On 10/13, Yunlong Song wrote: > > > This can help us to debug on some corner case. > > > > > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > > > Signed-off-by: Chao Yu <yuchao0@huawei.com> > > > --- > > > fs/f2fs/gc.c | 6 +++++- > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > > > index 197ebf4..2b03202 100644 > > > --- a/fs/f2fs/gc.c > > > +++ b/fs/f2fs/gc.c > > > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > .ilist = LIST_HEAD_INIT(gc_list.ilist), > > > .iroot = RADIX_TREE_INIT(GFP_NOFS), > > > }; > > > + bool need_fggc = false; > > > trace_f2fs_gc_begin(sbi->sb, sync, background, > > > get_pages(sbi, F2FS_DIRTY_NODES), > > > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > if (ret) > > > goto stop; > > > } > > > - if (has_not_enough_free_secs(sbi, 0, 0)) > > > + if (has_not_enough_free_secs(sbi, 0, 0)) { > > > gc_type = FG_GC; > > > + need_fggc = true; > > > + } > > > } > > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > > > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > goto stop; > > > } > > > if (!__get_victim(sbi, &segno, gc_type)) { > > > + f2fs_bug_on(sbi, !total_freed && need_fggc); > > Just like this? > That's OK. I'm not quite sure whether this is really a bug_on case. Let me make it WARN_ON() for debugging purpose first. Thanks, > > > > > f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); > > > > > ret = -ENODATA; > > > goto stop; > > > } > > > -- > > > 1.8.5.2 > > . > > > > -- > Thanks, > Yunlong Song > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-07 2:38 ` Jaegeuk Kim @ 2017-11-07 2:40 ` Jaegeuk Kim 2017-11-07 3:16 ` Yunlong Song 0 siblings, 1 reply; 27+ messages in thread From: Jaegeuk Kim @ 2017-11-07 2:40 UTC (permalink / raw) To: Yunlong Song Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 11/06, Jaegeuk Kim wrote: > On 11/06, Yunlong Song wrote: > > Agree. > > > > On 2017/11/3 11:44, Jaegeuk Kim wrote: > > > On 10/13, Yunlong Song wrote: > > > > This can help us to debug on some corner case. > > > > > > > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > > > > Signed-off-by: Chao Yu <yuchao0@huawei.com> > > > > --- > > > > fs/f2fs/gc.c | 6 +++++- > > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > > > > index 197ebf4..2b03202 100644 > > > > --- a/fs/f2fs/gc.c > > > > +++ b/fs/f2fs/gc.c > > > > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > .ilist = LIST_HEAD_INIT(gc_list.ilist), > > > > .iroot = RADIX_TREE_INIT(GFP_NOFS), > > > > }; > > > > + bool need_fggc = false; > > > > trace_f2fs_gc_begin(sbi->sb, sync, background, > > > > get_pages(sbi, F2FS_DIRTY_NODES), > > > > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > if (ret) > > > > goto stop; > > > > } > > > > - if (has_not_enough_free_secs(sbi, 0, 0)) > > > > + if (has_not_enough_free_secs(sbi, 0, 0)) { > > > > gc_type = FG_GC; > > > > + need_fggc = true; > > > > + } > > > > } > > > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > > > > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > goto stop; > > > > } > > > > if (!__get_victim(sbi, &segno, gc_type)) { > > > > + f2fs_bug_on(sbi, !total_freed && need_fggc); > > > Just like this? > > That's OK. > > I'm not quite sure whether this is really a bug_on case. > Let me make it WARN_ON() for debugging purpose first. BTW, why is this the special case where BG_GC detects FG_GC? > > Thanks, > > > > > > > > > f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); > > > > > > > ret = -ENODATA; > > > > goto stop; > > > > } > > > > -- > > > > 1.8.5.2 > > > . > > > > > > > -- > > Thanks, > > Yunlong Song > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-07 2:40 ` Jaegeuk Kim @ 2017-11-07 3:16 ` Yunlong Song 2017-11-07 3:26 ` Jaegeuk Kim 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-11-07 3:16 UTC (permalink / raw) To: Jaegeuk Kim Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel Because I find that some out-of-free problem is caused by the failure of get victim target. For example, chao has pointed out that he has found out a bug when adding this bug_on last week. On 2017/11/7 10:40, Jaegeuk Kim wrote: > On 11/06, Jaegeuk Kim wrote: >> On 11/06, Yunlong Song wrote: >>> Agree. >>> >>> On 2017/11/3 11:44, Jaegeuk Kim wrote: >>>> On 10/13, Yunlong Song wrote: >>>>> This can help us to debug on some corner case. >>>>> >>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>> --- >>>>> fs/f2fs/gc.c | 6 +++++- >>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>> >>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>> index 197ebf4..2b03202 100644 >>>>> --- a/fs/f2fs/gc.c >>>>> +++ b/fs/f2fs/gc.c >>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>> }; >>>>> + bool need_fggc = false; >>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>> if (ret) >>>>> goto stop; >>>>> } >>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>> gc_type = FG_GC; >>>>> + need_fggc = true; >>>>> + } >>>>> } >>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>> goto stop; >>>>> } >>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>> Just like this? >>> That's OK. >> I'm not quite sure whether this is really a bug_on case. >> Let me make it WARN_ON() for debugging purpose first. > BTW, why is this the special case where BG_GC detects FG_GC? > >> Thanks, >> >>>> f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); >>>> >>>>> ret = -ENODATA; >>>>> goto stop; >>>>> } >>>>> -- >>>>> 1.8.5.2 >>>> . >>>> >>> -- >>> Thanks, >>> Yunlong Song >>> > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-07 3:16 ` Yunlong Song @ 2017-11-07 3:26 ` Jaegeuk Kim 2017-11-07 4:01 ` Yunlong Song 0 siblings, 1 reply; 27+ messages in thread From: Jaegeuk Kim @ 2017-11-07 3:26 UTC (permalink / raw) To: Yunlong Song Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 11/07, Yunlong Song wrote: > Because I find that some out-of-free problem is caused by the failure > of get victim target. For example, chao has pointed out that he has > found out a bug when adding this bug_on last week. That's NOT what I asked. Why not checking FG_GC all the time like this? f2fs_bug_on(sbi, !total_freed && gc_type == FG_GC); > > On 2017/11/7 10:40, Jaegeuk Kim wrote: > > On 11/06, Jaegeuk Kim wrote: > > > On 11/06, Yunlong Song wrote: > > > > Agree. > > > > > > > > On 2017/11/3 11:44, Jaegeuk Kim wrote: > > > > > On 10/13, Yunlong Song wrote: > > > > > > This can help us to debug on some corner case. > > > > > > > > > > > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > > > > > > Signed-off-by: Chao Yu <yuchao0@huawei.com> > > > > > > --- > > > > > > fs/f2fs/gc.c | 6 +++++- > > > > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > > > > > > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > > > > > > index 197ebf4..2b03202 100644 > > > > > > --- a/fs/f2fs/gc.c > > > > > > +++ b/fs/f2fs/gc.c > > > > > > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > > > .ilist = LIST_HEAD_INIT(gc_list.ilist), > > > > > > .iroot = RADIX_TREE_INIT(GFP_NOFS), > > > > > > }; > > > > > > + bool need_fggc = false; > > > > > > trace_f2fs_gc_begin(sbi->sb, sync, background, > > > > > > get_pages(sbi, F2FS_DIRTY_NODES), > > > > > > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > > > if (ret) > > > > > > goto stop; > > > > > > } > > > > > > - if (has_not_enough_free_secs(sbi, 0, 0)) > > > > > > + if (has_not_enough_free_secs(sbi, 0, 0)) { > > > > > > gc_type = FG_GC; > > > > > > + need_fggc = true; > > > > > > + } > > > > > > } > > > > > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > > > > > > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > > > goto stop; > > > > > > } > > > > > > if (!__get_victim(sbi, &segno, gc_type)) { > > > > > > + f2fs_bug_on(sbi, !total_freed && need_fggc); > > > > > Just like this? > > > > That's OK. > > > I'm not quite sure whether this is really a bug_on case. > > > Let me make it WARN_ON() for debugging purpose first. > > BTW, why is this the special case where BG_GC detects FG_GC? > > > > > Thanks, > > > > > > > > f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); > > > > > > > > > > > ret = -ENODATA; > > > > > > goto stop; > > > > > > } > > > > > > -- > > > > > > 1.8.5.2 > > > > > . > > > > > > > > > -- > > > > Thanks, > > > > Yunlong Song > > > > > > . > > > > -- > Thanks, > Yunlong Song > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-07 3:26 ` Jaegeuk Kim @ 2017-11-07 4:01 ` Yunlong Song 2017-11-07 6:56 ` Chao Yu 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-11-07 4:01 UTC (permalink / raw) To: Jaegeuk Kim Cc: chao, yuchao0, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel Sorry, misunderstanding, because I think when sync == true, FG_GC does not check has_not_enough_free_secs, so maybe it does not have to do any gc at all. For example, if there are 100 segments for f2fs, and 20 segments are full or valid blocks over fggc_threshold, then it is correct to fail in get victim. On 2017/11/7 11:26, Jaegeuk Kim wrote: > On 11/07, Yunlong Song wrote: >> Because I find that some out-of-free problem is caused by the failure >> of get victim target. For example, chao has pointed out that he has >> found out a bug when adding this bug_on last week. > That's NOT what I asked. Why not checking FG_GC all the time like this? > > f2fs_bug_on(sbi, !total_freed && gc_type == FG_GC); > >> On 2017/11/7 10:40, Jaegeuk Kim wrote: >>> On 11/06, Jaegeuk Kim wrote: >>>> On 11/06, Yunlong Song wrote: >>>>> Agree. >>>>> >>>>> On 2017/11/3 11:44, Jaegeuk Kim wrote: >>>>>> On 10/13, Yunlong Song wrote: >>>>>>> This can help us to debug on some corner case. >>>>>>> >>>>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>>>> --- >>>>>>> fs/f2fs/gc.c | 6 +++++- >>>>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>>>> >>>>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>>>> index 197ebf4..2b03202 100644 >>>>>>> --- a/fs/f2fs/gc.c >>>>>>> +++ b/fs/f2fs/gc.c >>>>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>>>> }; >>>>>>> + bool need_fggc = false; >>>>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>> if (ret) >>>>>>> goto stop; >>>>>>> } >>>>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>>>> gc_type = FG_GC; >>>>>>> + need_fggc = true; >>>>>>> + } >>>>>>> } >>>>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>> goto stop; >>>>>>> } >>>>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>>>> Just like this? >>>>> That's OK. >>>> I'm not quite sure whether this is really a bug_on case. >>>> Let me make it WARN_ON() for debugging purpose first. >>> BTW, why is this the special case where BG_GC detects FG_GC? >>> >>>> Thanks, >>>> >>>>>> f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); >>>>>> >>>>>>> ret = -ENODATA; >>>>>>> goto stop; >>>>>>> } >>>>>>> -- >>>>>>> 1.8.5.2 >>>>>> . >>>>>> >>>>> -- >>>>> Thanks, >>>>> Yunlong Song >>>>> >>> . >>> >> -- >> Thanks, >> Yunlong Song >> > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-07 4:01 ` Yunlong Song @ 2017-11-07 6:56 ` Chao Yu 2017-11-08 3:06 ` Yunlong Song 0 siblings, 1 reply; 27+ messages in thread From: Chao Yu @ 2017-11-07 6:56 UTC (permalink / raw) To: Yunlong Song, Jaegeuk Kim Cc: chao, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 2017/11/7 12:01, Yunlong Song wrote: > Sorry, misunderstanding, because I think when sync == true, FG_GC does not > check has_not_enough_free_secs, so maybe it does not have to do any gc > at all. > For example, if there are 100 segments for f2fs, and 20 segments are full or > valid blocks over fggc_threshold, then it is correct to fail in get victim. > > > On 2017/11/7 11:26, Jaegeuk Kim wrote: >> On 11/07, Yunlong Song wrote: >>> Because I find that some out-of-free problem is caused by the failure >>> of get victim target. For example, chao has pointed out that he has >>> found out a bug when adding this bug_on last week. >> That's NOT what I asked. Why not checking FG_GC all the time like this? >> >> f2fs_bug_on(sbi, !total_freed && gc_type == FG_GC); ioctl(F2FS_IOC_GARBAGE_COLLECT, &1) will simply trigger this bug_on, so we have to check the conditon only when we run out-of-free-space? Thanks, >> >>> On 2017/11/7 10:40, Jaegeuk Kim wrote: >>>> On 11/06, Jaegeuk Kim wrote: >>>>> On 11/06, Yunlong Song wrote: >>>>>> Agree. >>>>>> >>>>>> On 2017/11/3 11:44, Jaegeuk Kim wrote: >>>>>>> On 10/13, Yunlong Song wrote: >>>>>>>> This can help us to debug on some corner case. >>>>>>>> >>>>>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>>>>> --- >>>>>>>> fs/f2fs/gc.c | 6 +++++- >>>>>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>>>>> >>>>>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>>>>> index 197ebf4..2b03202 100644 >>>>>>>> --- a/fs/f2fs/gc.c >>>>>>>> +++ b/fs/f2fs/gc.c >>>>>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>>>>> }; >>>>>>>> + bool need_fggc = false; >>>>>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>> if (ret) >>>>>>>> goto stop; >>>>>>>> } >>>>>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>>>>> gc_type = FG_GC; >>>>>>>> + need_fggc = true; >>>>>>>> + } >>>>>>>> } >>>>>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>> goto stop; >>>>>>>> } >>>>>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>>>>> Just like this? >>>>>> That's OK. >>>>> I'm not quite sure whether this is really a bug_on case. >>>>> Let me make it WARN_ON() for debugging purpose first. >>>> BTW, why is this the special case where BG_GC detects FG_GC? >>>> >>>>> Thanks, >>>>> >>>>>>> f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); >>>>>>> >>>>>>>> ret = -ENODATA; >>>>>>>> goto stop; >>>>>>>> } >>>>>>>> -- >>>>>>>> 1.8.5.2 >>>>>>> . >>>>>>> >>>>>> -- >>>>>> Thanks, >>>>>> Yunlong Song >>>>>> >>>> . >>>> >>> -- >>> Thanks, >>> Yunlong Song >>> >> . >> > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-07 6:56 ` Chao Yu @ 2017-11-08 3:06 ` Yunlong Song 2017-11-09 17:59 ` Jaegeuk Kim 0 siblings, 1 reply; 27+ messages in thread From: Yunlong Song @ 2017-11-08 3:06 UTC (permalink / raw) To: Chao Yu, Jaegeuk Kim Cc: chao, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel So we should use f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); On 2017/11/7 14:56, Chao Yu wrote: > On 2017/11/7 12:01, Yunlong Song wrote: >> Sorry, misunderstanding, because I think when sync == true, FG_GC does not >> check has_not_enough_free_secs, so maybe it does not have to do any gc >> at all. >> For example, if there are 100 segments for f2fs, and 20 segments are full or >> valid blocks over fggc_threshold, then it is correct to fail in get victim. >> >> >> On 2017/11/7 11:26, Jaegeuk Kim wrote: >>> On 11/07, Yunlong Song wrote: >>>> Because I find that some out-of-free problem is caused by the failure >>>> of get victim target. For example, chao has pointed out that he has >>>> found out a bug when adding this bug_on last week. >>> That's NOT what I asked. Why not checking FG_GC all the time like this? >>> >>> f2fs_bug_on(sbi, !total_freed && gc_type == FG_GC); > ioctl(F2FS_IOC_GARBAGE_COLLECT, &1) will simply trigger this bug_on, so we > have to check the conditon only when we run out-of-free-space? > > Thanks, > >>>> On 2017/11/7 10:40, Jaegeuk Kim wrote: >>>>> On 11/06, Jaegeuk Kim wrote: >>>>>> On 11/06, Yunlong Song wrote: >>>>>>> Agree. >>>>>>> >>>>>>> On 2017/11/3 11:44, Jaegeuk Kim wrote: >>>>>>>> On 10/13, Yunlong Song wrote: >>>>>>>>> This can help us to debug on some corner case. >>>>>>>>> >>>>>>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>>>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>>>>>> --- >>>>>>>>> fs/f2fs/gc.c | 6 +++++- >>>>>>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>>>>>> >>>>>>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>>>>>> index 197ebf4..2b03202 100644 >>>>>>>>> --- a/fs/f2fs/gc.c >>>>>>>>> +++ b/fs/f2fs/gc.c >>>>>>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>>>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>>>>>> }; >>>>>>>>> + bool need_fggc = false; >>>>>>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>>>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>>>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>>> if (ret) >>>>>>>>> goto stop; >>>>>>>>> } >>>>>>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>>>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>>>>>> gc_type = FG_GC; >>>>>>>>> + need_fggc = true; >>>>>>>>> + } >>>>>>>>> } >>>>>>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>>>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>>> goto stop; >>>>>>>>> } >>>>>>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>>>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>>>>>> Just like this? >>>>>>> That's OK. >>>>>> I'm not quite sure whether this is really a bug_on case. >>>>>> Let me make it WARN_ON() for debugging purpose first. >>>>> BTW, why is this the special case where BG_GC detects FG_GC? >>>>> >>>>>> Thanks, >>>>>> >>>>>>>> f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); >>>>>>>> >>>>>>>>> ret = -ENODATA; >>>>>>>>> goto stop; >>>>>>>>> } >>>>>>>>> -- >>>>>>>>> 1.8.5.2 >>>>>>>> . >>>>>>>> >>>>>>> -- >>>>>>> Thanks, >>>>>>> Yunlong Song >>>>>>> >>>>> . >>>>> >>>> -- >>>> Thanks, >>>> Yunlong Song >>>> >>> . >>> > > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-08 3:06 ` Yunlong Song @ 2017-11-09 17:59 ` Jaegeuk Kim 2017-11-10 2:11 ` Yunlong Song 0 siblings, 1 reply; 27+ messages in thread From: Jaegeuk Kim @ 2017-11-09 17:59 UTC (permalink / raw) To: Yunlong Song Cc: Chao Yu, chao, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel On 11/08, Yunlong Song wrote: > So we should use f2fs_bug_on(sbi, !total_freed && !sync && gc_type == > FG_GC); f2fs_bug_on(sbi, !total_freed && has_not_enough_free_secs(sbi, 0, 0)); ? > > On 2017/11/7 14:56, Chao Yu wrote: > > On 2017/11/7 12:01, Yunlong Song wrote: > > > Sorry, misunderstanding, because I think when sync == true, FG_GC does not > > > check has_not_enough_free_secs, so maybe it does not have to do any gc > > > at all. > > > For example, if there are 100 segments for f2fs, and 20 segments are full or > > > valid blocks over fggc_threshold, then it is correct to fail in get victim. > > > > > > > > > On 2017/11/7 11:26, Jaegeuk Kim wrote: > > > > On 11/07, Yunlong Song wrote: > > > > > Because I find that some out-of-free problem is caused by the failure > > > > > of get victim target. For example, chao has pointed out that he has > > > > > found out a bug when adding this bug_on last week. > > > > That's NOT what I asked. Why not checking FG_GC all the time like this? > > > > > > > > f2fs_bug_on(sbi, !total_freed && gc_type == FG_GC); > > ioctl(F2FS_IOC_GARBAGE_COLLECT, &1) will simply trigger this bug_on, so we > > have to check the conditon only when we run out-of-free-space? > > > > Thanks, > > > > > > > On 2017/11/7 10:40, Jaegeuk Kim wrote: > > > > > > On 11/06, Jaegeuk Kim wrote: > > > > > > > On 11/06, Yunlong Song wrote: > > > > > > > > Agree. > > > > > > > > > > > > > > > > On 2017/11/3 11:44, Jaegeuk Kim wrote: > > > > > > > > > On 10/13, Yunlong Song wrote: > > > > > > > > > > This can help us to debug on some corner case. > > > > > > > > > > > > > > > > > > > > Signed-off-by: Yunlong Song <yunlong.song@huawei.com> > > > > > > > > > > Signed-off-by: Chao Yu <yuchao0@huawei.com> > > > > > > > > > > --- > > > > > > > > > > fs/f2fs/gc.c | 6 +++++- > > > > > > > > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > > > > > > > > > > > > > > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > > > > > > > > > > index 197ebf4..2b03202 100644 > > > > > > > > > > --- a/fs/f2fs/gc.c > > > > > > > > > > +++ b/fs/f2fs/gc.c > > > > > > > > > > @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > > > > > > > .ilist = LIST_HEAD_INIT(gc_list.ilist), > > > > > > > > > > .iroot = RADIX_TREE_INIT(GFP_NOFS), > > > > > > > > > > }; > > > > > > > > > > + bool need_fggc = false; > > > > > > > > > > trace_f2fs_gc_begin(sbi->sb, sync, background, > > > > > > > > > > get_pages(sbi, F2FS_DIRTY_NODES), > > > > > > > > > > @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > > > > > > > if (ret) > > > > > > > > > > goto stop; > > > > > > > > > > } > > > > > > > > > > - if (has_not_enough_free_secs(sbi, 0, 0)) > > > > > > > > > > + if (has_not_enough_free_secs(sbi, 0, 0)) { > > > > > > > > > > gc_type = FG_GC; > > > > > > > > > > + need_fggc = true; > > > > > > > > > > + } > > > > > > > > > > } > > > > > > > > > > /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ > > > > > > > > > > @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, > > > > > > > > > > goto stop; > > > > > > > > > > } > > > > > > > > > > if (!__get_victim(sbi, &segno, gc_type)) { > > > > > > > > > > + f2fs_bug_on(sbi, !total_freed && need_fggc); > > > > > > > > > Just like this? > > > > > > > > That's OK. > > > > > > > I'm not quite sure whether this is really a bug_on case. > > > > > > > Let me make it WARN_ON() for debugging purpose first. > > > > > > BTW, why is this the special case where BG_GC detects FG_GC? > > > > > > > > > > > > > Thanks, > > > > > > > > > > > > > > > > f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); > > > > > > > > > > > > > > > > > > > ret = -ENODATA; > > > > > > > > > > goto stop; > > > > > > > > > > } > > > > > > > > > > -- > > > > > > > > > > 1.8.5.2 > > > > > > > > > . > > > > > > > > > > > > > > > > > -- > > > > > > > > Thanks, > > > > > > > > Yunlong Song > > > > > > > > > > > > > > . > > > > > > > > > > > -- > > > > > Thanks, > > > > > Yunlong Song > > > > > > > > > . > > > > > > > > . > > > > -- > Thanks, > Yunlong Song > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v2] f2fs: add bug_on when f2fs_gc even fails to get one victim 2017-11-09 17:59 ` Jaegeuk Kim @ 2017-11-10 2:11 ` Yunlong Song 0 siblings, 0 replies; 27+ messages in thread From: Yunlong Song @ 2017-11-10 2:11 UTC (permalink / raw) To: Jaegeuk Kim Cc: Chao Yu, chao, yunlong.song, miaoxie, bintian.wang, linux-fsdevel, linux-f2fs-devel, linux-kernel Agree. On 2017/11/10 1:59, Jaegeuk Kim wrote: > On 11/08, Yunlong Song wrote: >> So we should use f2fs_bug_on(sbi, !total_freed && !sync && gc_type == >> FG_GC); > f2fs_bug_on(sbi, !total_freed && has_not_enough_free_secs(sbi, 0, 0)); ? > >> On 2017/11/7 14:56, Chao Yu wrote: >>> On 2017/11/7 12:01, Yunlong Song wrote: >>>> Sorry, misunderstanding, because I think when sync == true, FG_GC does not >>>> check has_not_enough_free_secs, so maybe it does not have to do any gc >>>> at all. >>>> For example, if there are 100 segments for f2fs, and 20 segments are full or >>>> valid blocks over fggc_threshold, then it is correct to fail in get victim. >>>> >>>> >>>> On 2017/11/7 11:26, Jaegeuk Kim wrote: >>>>> On 11/07, Yunlong Song wrote: >>>>>> Because I find that some out-of-free problem is caused by the failure >>>>>> of get victim target. For example, chao has pointed out that he has >>>>>> found out a bug when adding this bug_on last week. >>>>> That's NOT what I asked. Why not checking FG_GC all the time like this? >>>>> >>>>> f2fs_bug_on(sbi, !total_freed && gc_type == FG_GC); >>> ioctl(F2FS_IOC_GARBAGE_COLLECT, &1) will simply trigger this bug_on, so we >>> have to check the conditon only when we run out-of-free-space? >>> >>> Thanks, >>> >>>>>> On 2017/11/7 10:40, Jaegeuk Kim wrote: >>>>>>> On 11/06, Jaegeuk Kim wrote: >>>>>>>> On 11/06, Yunlong Song wrote: >>>>>>>>> Agree. >>>>>>>>> >>>>>>>>> On 2017/11/3 11:44, Jaegeuk Kim wrote: >>>>>>>>>> On 10/13, Yunlong Song wrote: >>>>>>>>>>> This can help us to debug on some corner case. >>>>>>>>>>> >>>>>>>>>>> Signed-off-by: Yunlong Song <yunlong.song@huawei.com> >>>>>>>>>>> Signed-off-by: Chao Yu <yuchao0@huawei.com> >>>>>>>>>>> --- >>>>>>>>>>> fs/f2fs/gc.c | 6 +++++- >>>>>>>>>>> 1 file changed, 5 insertions(+), 1 deletion(-) >>>>>>>>>>> >>>>>>>>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c >>>>>>>>>>> index 197ebf4..2b03202 100644 >>>>>>>>>>> --- a/fs/f2fs/gc.c >>>>>>>>>>> +++ b/fs/f2fs/gc.c >>>>>>>>>>> @@ -986,6 +986,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>>>>> .ilist = LIST_HEAD_INIT(gc_list.ilist), >>>>>>>>>>> .iroot = RADIX_TREE_INIT(GFP_NOFS), >>>>>>>>>>> }; >>>>>>>>>>> + bool need_fggc = false; >>>>>>>>>>> trace_f2fs_gc_begin(sbi->sb, sync, background, >>>>>>>>>>> get_pages(sbi, F2FS_DIRTY_NODES), >>>>>>>>>>> @@ -1018,8 +1019,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>>>>> if (ret) >>>>>>>>>>> goto stop; >>>>>>>>>>> } >>>>>>>>>>> - if (has_not_enough_free_secs(sbi, 0, 0)) >>>>>>>>>>> + if (has_not_enough_free_secs(sbi, 0, 0)) { >>>>>>>>>>> gc_type = FG_GC; >>>>>>>>>>> + need_fggc = true; >>>>>>>>>>> + } >>>>>>>>>>> } >>>>>>>>>>> /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ >>>>>>>>>>> @@ -1028,6 +1031,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, >>>>>>>>>>> goto stop; >>>>>>>>>>> } >>>>>>>>>>> if (!__get_victim(sbi, &segno, gc_type)) { >>>>>>>>>>> + f2fs_bug_on(sbi, !total_freed && need_fggc); >>>>>>>>>> Just like this? >>>>>>>>> That's OK. >>>>>>>> I'm not quite sure whether this is really a bug_on case. >>>>>>>> Let me make it WARN_ON() for debugging purpose first. >>>>>>> BTW, why is this the special case where BG_GC detects FG_GC? >>>>>>> >>>>>>>> Thanks, >>>>>>>> >>>>>>>>>> f2fs_bug_on(sbi, !total_freed && !sync && gc_type == FG_GC); >>>>>>>>>> >>>>>>>>>>> ret = -ENODATA; >>>>>>>>>>> goto stop; >>>>>>>>>>> } >>>>>>>>>>> -- >>>>>>>>>>> 1.8.5.2 >>>>>>>>>> . >>>>>>>>>> >>>>>>>>> -- >>>>>>>>> Thanks, >>>>>>>>> Yunlong Song >>>>>>>>> >>>>>>> . >>>>>>> >>>>>> -- >>>>>> Thanks, >>>>>> Yunlong Song >>>>>> >>>>> . >>>>> >>> . >>> >> -- >> Thanks, >> Yunlong Song >> > . > -- Thanks, Yunlong Song ^ permalink raw reply [flat|nested] 27+ messages in thread
end of thread, other threads:[~2017-11-10 2:13 UTC | newest] Thread overview: 27+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2017-10-11 13:51 [PATCH] f2fs: add bug_on when f2fs_gc even fails to get one victim Yunlong Song 2017-10-12 23:23 ` Jaegeuk Kim 2017-10-13 1:09 ` Yunlong Song 2017-10-13 2:29 ` Jaegeuk Kim 2017-10-13 11:09 ` [f2fs-dev] " Chao Yu 2017-10-13 11:20 ` Yunlong Song 2017-10-13 12:24 ` Chao Yu 2017-10-13 13:31 ` [PATCH v2] " Yunlong Song 2017-10-14 0:17 ` [f2fs-dev] " Chao Yu 2017-10-14 12:34 ` Yunlong Song 2017-10-16 3:25 ` Chao Yu 2017-10-31 1:32 ` Yunlong Song 2017-10-31 3:17 ` Chao Yu 2017-10-31 3:30 ` Yunlong Song 2017-11-03 3:28 ` Yunlong Song 2017-11-03 3:44 ` Jaegeuk Kim 2017-11-03 15:08 ` 宋云龙 2017-11-06 1:12 ` Yunlong Song 2017-11-07 2:38 ` Jaegeuk Kim 2017-11-07 2:40 ` Jaegeuk Kim 2017-11-07 3:16 ` Yunlong Song 2017-11-07 3:26 ` Jaegeuk Kim 2017-11-07 4:01 ` Yunlong Song 2017-11-07 6:56 ` Chao Yu 2017-11-08 3:06 ` Yunlong Song 2017-11-09 17:59 ` Jaegeuk Kim 2017-11-10 2:11 ` Yunlong Song
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).