* [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items
@ 2025-07-25 3:40 Qu Wenruo
2025-07-25 3:40 ` [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly Qu Wenruo
` (3 more replies)
0 siblings, 4 replies; 7+ messages in thread
From: Qu Wenruo @ 2025-07-25 3:40 UTC (permalink / raw)
To: linux-btrfs
There is an internal bug report that some half-dropped subvolume makes
balance to fail.
It turns out that, the involved subvolume doesn't have an orphan item,
this means the half-dropped subvolume is never going to be cleaned up.
Then at balance time, a reloc tree is created for that half-dropped
subvolume, and since balance doesn't expect to get a half-dropped
subvolume, it doesn't check the drop_process_key and increased ref on
already dropped nodes.
The problem for progs is that, neither original mode nor lowmem detects
this kind of problem.
Original mode has a bad logic which prevents us from calling the proper
orphan item check.
Lowmem mode doesn't even take orphan item into consideration.
This series will add the detection part for btrfs-check, for both
original and lowmem mode, with a hand crafted image for test.
Qu Wenruo (3):
btrfs-progs: check/original: detect missing orphan items correctly
btrfs-progs: check/lowmem: detect missing orphan items correctly
btrfs-progs: fsck-tests: a new test case for missing root orphan item
check/main.c | 3 +--
check/mode-lowmem.c | 11 +++++++++++
.../066-missing-root-orphan-item/default.img.xz | Bin 0 -> 13468 bytes
.../066-missing-root-orphan-item/test.sh | 14 ++++++++++++++
4 files changed, 26 insertions(+), 2 deletions(-)
create mode 100644 tests/fsck-tests/066-missing-root-orphan-item/default.img.xz
create mode 100755 tests/fsck-tests/066-missing-root-orphan-item/test.sh
--
2.50.0
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly
2025-07-25 3:40 [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Qu Wenruo
@ 2025-07-25 3:40 ` Qu Wenruo
2025-07-25 18:09 ` Boris Burkov
2025-07-25 3:40 ` [PATCH 2/3] btrfs-progs: check/lowmem: " Qu Wenruo
` (2 subsequent siblings)
3 siblings, 1 reply; 7+ messages in thread
From: Qu Wenruo @ 2025-07-25 3:40 UTC (permalink / raw)
To: linux-btrfs
[BUG]
If we have a filesystem with a subvolume that has 0 refs but without an
orphan item, btrfs check won't report any error on it.
$ btrfs ins dump-tree -t root /dev/test/scratch1
btrfs-progs v6.15
root tree
node 5242880 level 1 items 2 free space 119 generation 11 owner ROOT_TREE
node 5242880 flags 0x1(WRITTEN) backref revision 1
fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
key (EXTENT_TREE ROOT_ITEM 0) block 5267456 gen 11
key (ROOT_TREE_DIR DIR_ITEM 2378154706) block 5246976 gen 11
leaf 5267456 items 6 free space 2339 generation 11 owner ROOT_TREE
leaf 5267456 flags 0x1(WRITTEN) backref revision 1
fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
[...]
leaf 5246976 items 6 free space 1613 generation 11 owner ROOT_TREE
leaf 5246976 flags 0x1(WRITTEN) backref revision 1
checksum stored 47620783
checksum calced 47620783
fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
[...]
item 4 key (256 ROOT_ITEM 0) itemoff 2202 itemsize 439
generation 9 root_dirid 256 bytenr 5898240 byte_limit 0 bytes_used 581632
last_snapshot 0 flags 0x1000000000000(none) refs 0 <<<
drop_progress key (0 UNKNOWN.0 0) drop_level 0
level 2 generation_v2 9
item 5 key (DATA_RELOC_TREE ROOT_ITEM 0) itemoff 1763 itemsize 439
generation 5 root_dirid 256 bytenr 5287936 byte_limit 0 bytes_used 4096
last_snapshot 0 flags 0x0(none) refs 1
drop_progress key (0 UNKNOWN.0 0) drop_level 0
level 0 generation_v2 5
^^^ No orphan item for subvolume 256.
Then "btrfs check" will not report anything wrong with it:
Opening filesystem to check...
Checking filesystem on /dev/test/scratch1
UUID: ff32309e-4e90-4402-b1ef-0a1f9f28bfff
[1/8] checking log skipped (none written)
[2/8] checking root items
[3/8] checking extents
[4/8] checking free space tree
[5/8] checking fs roots
[6/8] checking only csums items (without verifying data)
[7/8] checking root refs
[8/8] checking quota groups skipped (not enabled on this FS)
found 638976 bytes used, no error found <<<
total csum bytes: 0
total tree bytes: 638976
total fs tree bytes: 589824
total extent tree bytes: 16384
btree space waste bytes: 289501
file data blocks allocated: 0
referenced 0
[CAUSE]
Although we have check_orphan_item() call inside check_root_refs(), it
relies the root record to have its 'found_root_item' member set.
Otherwise it will not report this as an error.
But that 'found_root_item' is always set to 0, if the subvolume has zero
ref, this means any subvolume with 0 refs will not have its orphan item
checked.
[FIX]
Set root_record::found_root_item to 1 inside check_fs_root().
As check_fs_root() is called after we found a root item, we should set
root_record::found_root_item to indicate this fact, and allows proper
orphan item check to be done.
Signed-off-by: Qu Wenruo <wqu@suse.com>
---
check/main.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/check/main.c b/check/main.c
index b78eb59d0c50..1536c1bbbccd 100644
--- a/check/main.c
+++ b/check/main.c
@@ -3728,8 +3728,7 @@ static int check_fs_root(struct btrfs_root *root,
if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
rec = get_root_rec(root_cache, root->root_key.objectid);
BUG_ON(IS_ERR(rec));
- if (btrfs_root_refs(root_item) > 0)
- rec->found_root_item = 1;
+ rec->found_root_item = 1;
}
memset(&root_node, 0, sizeof(root_node));
--
2.50.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH 2/3] btrfs-progs: check/lowmem: detect missing orphan items correctly
2025-07-25 3:40 [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Qu Wenruo
2025-07-25 3:40 ` [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly Qu Wenruo
@ 2025-07-25 3:40 ` Qu Wenruo
2025-07-25 3:40 ` [PATCH 3/3] btrfs-progs: fsck-tests: a new test case for missing root orphan item Qu Wenruo
2025-07-25 18:10 ` [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Boris Burkov
3 siblings, 0 replies; 7+ messages in thread
From: Qu Wenruo @ 2025-07-25 3:40 UTC (permalink / raw)
To: linux-btrfs
[BUG]
If we have a filesystem with a subvolume that has 0 refs but without an
orphan item, btrfs check won't report any error on it.
$ btrfs ins dump-tree -t root /dev/test/scratch1
btrfs-progs v6.15
root tree
node 5242880 level 1 items 2 free space 119 generation 11 owner ROOT_TREE
node 5242880 flags 0x1(WRITTEN) backref revision 1
fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
key (EXTENT_TREE ROOT_ITEM 0) block 5267456 gen 11
key (ROOT_TREE_DIR DIR_ITEM 2378154706) block 5246976 gen 11
leaf 5267456 items 6 free space 2339 generation 11 owner ROOT_TREE
leaf 5267456 flags 0x1(WRITTEN) backref revision 1
fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
[...]
leaf 5246976 items 6 free space 1613 generation 11 owner ROOT_TREE
leaf 5246976 flags 0x1(WRITTEN) backref revision 1
checksum stored 47620783
checksum calced 47620783
fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
[...]
item 4 key (256 ROOT_ITEM 0) itemoff 2202 itemsize 439
generation 9 root_dirid 256 bytenr 5898240 byte_limit 0 bytes_used 581632
last_snapshot 0 flags 0x1000000000000(none) refs 0 <<<
drop_progress key (0 UNKNOWN.0 0) drop_level 0
level 2 generation_v2 9
item 5 key (DATA_RELOC_TREE ROOT_ITEM 0) itemoff 1763 itemsize 439
generation 5 root_dirid 256 bytenr 5287936 byte_limit 0 bytes_used 4096
last_snapshot 0 flags 0x0(none) refs 1
drop_progress key (0 UNKNOWN.0 0) drop_level 0
level 0 generation_v2 5
^^^ No orphan item for subvolume 256.
Then "btrfs check --mode=lowmem" will not report anything wrong with it:
Opening filesystem to check...
Checking filesystem on /dev/test/scratch1
UUID: ff32309e-4e90-4402-b1ef-0a1f9f28bfff
[1/8] checking log skipped (none written)
[2/8] checking root items
[3/8] checking extents
[4/8] checking free space tree
[5/8] checking fs roots
[6/8] checking only csums items (without verifying data)
[7/8] checking root refs done with fs roots in lowmem mode, skipping
[8/8] checking quota groups skipped (not enabled on this FS)
found 638976 bytes used, no error found <<
total csum bytes: 0
total tree bytes: 638976
total fs tree bytes: 589824
total extent tree bytes: 16384
btree space waste bytes: 289501
file data blocks allocated: 0
referenced 0
[CAUSE]
In lowmem's check_btrfs_root() we have no special handling for refs 0,
as long as there is no ROOT_BACKREF for the orphan root, everything will
be reported as OK.
But we should expect an orphan item for an orphan subvolume, this means
lowmem doesn't do any checks against the orphan item for a subvolume.
[FIX]
For subvolume trees, if its refs is 0, also check if we have an orphan
item for it.
If not, report this as an error.
Signed-off-by: Qu Wenruo <wqu@suse.com>
---
check/mode-lowmem.c | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/check/mode-lowmem.c b/check/mode-lowmem.c
index 5291070df025..e05bb0da1709 100644
--- a/check/mode-lowmem.c
+++ b/check/mode-lowmem.c
@@ -5502,6 +5502,17 @@ static int check_btrfs_root(struct btrfs_root *root, int check_all)
}
}
}
+ /*
+ * If this tree is a subvolume (not a reloc tree) and has no refs, there
+ * should be an orphan item for it, or this subvolume will never be deleted.
+ */
+ if (btrfs_root_refs(root_item) == 0 && is_fstree(btrfs_root_id(root))) {
+ if (!has_orphan_item(root->fs_info->tree_root,
+ btrfs_root_id(root))) {
+ error("missing orphan item for root %lld", btrfs_root_id(root));
+ err |= REFERENCER_MISSING;
+ }
+ }
if (btrfs_root_refs(root_item) > 0 ||
btrfs_disk_key_objectid(&root_item->drop_progress) == 0) {
path.nodes[level] = root->node;
--
2.50.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH 3/3] btrfs-progs: fsck-tests: a new test case for missing root orphan item
2025-07-25 3:40 [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Qu Wenruo
2025-07-25 3:40 ` [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly Qu Wenruo
2025-07-25 3:40 ` [PATCH 2/3] btrfs-progs: check/lowmem: " Qu Wenruo
@ 2025-07-25 3:40 ` Qu Wenruo
2025-07-25 18:10 ` [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Boris Burkov
3 siblings, 0 replies; 7+ messages in thread
From: Qu Wenruo @ 2025-07-25 3:40 UTC (permalink / raw)
To: linux-btrfs
The new image is created by:
- Make btrfs_delete_subvolume() to commit transaction
So that we're sure the orphan item is committed to disk after removing
the subvolume.
- Remove the orphan item from tree root
Then use this image to make sure btrfs check can detect such problem.
Repair is not yet supported.
Signed-off-by: Qu Wenruo <wqu@suse.com>
---
.../066-missing-root-orphan-item/default.img.xz | Bin 0 -> 13468 bytes
.../066-missing-root-orphan-item/test.sh | 14 ++++++++++++++
2 files changed, 14 insertions(+)
create mode 100644 tests/fsck-tests/066-missing-root-orphan-item/default.img.xz
create mode 100755 tests/fsck-tests/066-missing-root-orphan-item/test.sh
diff --git a/tests/fsck-tests/066-missing-root-orphan-item/default.img.xz b/tests/fsck-tests/066-missing-root-orphan-item/default.img.xz
new file mode 100644
index 0000000000000000000000000000000000000000..c34f989daf0aef34e2692f2d11d4d4b1417ccdb5
GIT binary patch
literal 13468
zcmexsUKJ6=z`*kC+7>sK1Gh378W<HB#TdYVqqN!jf&G7zpjZainSIM6_ia9SQmQh?
zJiPVUv7OJaD(*WJdfw&W?$y7Z&u=;{UlQb0zu>^Vzu6D=I-CoTY>?OcAi}1*^V!nK
zsJfSjc0Iljy#Ls}qz96l!`mkW|2VMVRJ(fA);;?e^0^kXJQl9iv8{^!dZNnWy6Qti
z_4`3>3cq=jPI7MV%5kui|1Ko)^YF3+p*D?_ckd;+YwGrV^}HL?bs)*BobTh_()BXJ
zUze~H2rhXsskb>-?D^J~v|Dc{EdMFVma|>RZkC+%!HMTA?XS&gO#GC%<!N%>+M~)#
zomqM3)a~+~Sn)LK_LZeoJI_7sFyA^sNqx=ZV?s+k*3@fNiYBjJujp7KT3dFqJe*x%
z#>|<~dtSH5tjOIr$D8lfrC<e(j15LV>ueG)26Q}m#P+mi^Uw6)(;wzZZT>G77kF#C
z(cAiO|K0~EKGjfC+?ce2YhJdF>a{(wTQyfoPN<Umzu+kEzO8aoB$&1bE9NfRtU5b!
zo{&LScj&=;bI;bVUle%!?R}=Sgob6wVWmuyE(#RHWk&bK_+R>C+OeTf?0!t&3+~oj
zU0oIJa}#y*PiU!2b}g6w`keg$<DH~wz21`ZGkLQP=DFIgSvPHh&Z+YoO7_`*&3?}O
zVMgJJQ(J=rqj&##c`ag@AWNvvruTVSzh+(Clf0N|`rVhy4)gr=z4ltn(Dv$uxdnTz
z{2U^f7{%50ecyC;hHus~vHS(zyvq}h+uXK&@Gm0Zrgha``68|KBb&QaKW#g|;9`zk
zE$e*aGl^^R*^`961T4wl#rU);)n4%L)4K|X94!`zUV7Wv`t{`8U6Gvq(}NGmt&g1f
za2`+nml^M4Y8od0I6IlgQEi54{pvY=>#r_8d9ughr<Nwi=WWUsMbGu_*VKC%>BecR
z&Qd@B=kJREkM11*^r~r#taWa*R_<D7_RD)s?E$XiPaZ8Q*VS7Z_jR@Z;!LCc{q0}2
zy$NjYUJ>QJ<9$e7ZFH2S$?er0{I}FUzTC%ZW3_ui?D1Eho(Jua-y3`UK%$17@Y;|6
znI3C77P5S}eebToN9TEqRE-~7)_o~VQ8KYU9C$GA(&df0{WSq>OEcu$CDvp`XziUm
z-I{0Wi6&c~+fDJEE4xoz6iVJ5?&<l(;1}z!{NqorXJ7LE%NFQ;pKZHD(8<@V7E6A6
zsLL*@Nu2WSXq+r-#xf?hH$BJqPTaeoG-k$;=Wfdbr%%hNK6ZEEgEMcql>-8A37_Sa
zlWpQVA@yy_jpVyF&F`1|Q;>AXW!gV`=0i>U#IqB8jN{+M&(<|r_D{sU{qMf04QpQ|
z_)hNG6nOdnm4BN&Cas^htm>iUF{yd84eXB3zPapFsD*;WtS`nlPfQAtu@(FKz;CbU
zlxKbM`=U7AWjMO*otmT0x~qI<@ObGLQ`5*VaqjfaEmsOBmM(J?S?jyr;V1inx2g+x
z8kX?hmt#IRqx;~Zzlt{;Q-T`aE%7>hiOt>kK+WIp@#lB!`eJ6gMXu+<zq{v`E1dUF
zSUvS}ZcuvGgso-QT#7W4V~h_Oca{|$GCg$7Q6@n!*Pyxi{!jPq-L9GD{L&Q>w|X`n
zly($|%(@zuvovIG=FZwDQsNp#dM7TexpFLcd$!CmeSUlU+X??`mC~J96eY4)3foJx
z(`2*b=PhTAQk)xZ|1ahfNA!I*zBS+Kq?dn*IjS7<PF~jK`5SSk!$;S=xWwpVr6w8w
zwCuryAMyQDBXyaQ56pTj#yI=l*GTT@eX2r+%~P@_>B*@3sRrf5=G^sHYO$|r**DMb
z;hpXos}`o*6_`HFuqXOS%)_HS6IxCL{dgb}!?j?^f|qM`f7#xeB$l(G`lyb{!9`bJ
zg&kq4Qq|MiRyQR*<!0YHxkIj^HE-I}IxV*+HVIFQ_6b~b{L4k5x)1VS8m~w0-dXiJ
zYu1hbd5`X%`hJ@8>K38y*WuIh4|A6s$&ym#X`A0X%VsbCuaIZD=WeNddCu^Nv%oO?
z*e_8B1yBE5%jRA?8mP{;#W2n3>h+?D51(33oWpn3?(lih&EJ?A&hJqw+*7=9{@qzC
zBipC2f4lo&O46nOmaEeXmgn60e(r8ckzbyu+!iaxNz*5v_u27l{+lV{`zLwkJZ3b!
z7Gt&YY)AQmGs}O!$=aN<;I!(Dd;U+F7OY-;EJF0M-zWVEx7kz~<5pfjbJy+h8yy|1
z{6B}XuAk9yxOwNdq21^5h4J&gm;}ylZd2WPNn~HxzbmI+^7nN`TOVP%q4!90{z=D~
zS0Y7zPP*D1tsTf`r9RE=(Dl9d59HtOY~v3XQu(NvwByI}9I1lJ4n^M$f4E9!o#<GS
zvD);)lbPYuix+&^b&|d6>Dfc7MG0EBm{Y$>)Onsip&`Czb*uK8L@A~HbA{Dr{J3&#
z)}kq@A>VDUW@Y^>XWXWik>B%z?bdX)UCl3R{I12e#4o)mpd^>=KXHAkN<!kpHK8tl
zkNww3y!n^)%J0gvduCV&WtT30o%c>fVvB7=WTcnzxBK%P*xGs{Sx*bw-^tKC#PQu=
zvzZvrFWI&TsW_PmooglMiq2NtFPUQ29{4}ovLSWl(u@C1E_GK}J`wuacdYk^<na@%
zL5UNtJ-=r7bpL_>hm~5?7yPn+x-{ic@ri}0u4hC{lYd-0V6U+8M`*~2iy^J*anVO_
znAfyK<eqsj@0@q@ao5NCpXY5l|7uU?30CX36Wg~PH99=CX8)CE6J75!I<$ygeV;p7
zy;!)%q0}M1cH3*s4z1M!pHBAXbafZ>iJtrTi+`?hKp1yhmI2R?2@YR%ee~t$o-5v*
zFaLz2P{8>h&!p#P-WR^^{QuR$x^>n19i0E%&M0-A-64FrB&?-sPM)B|<Im3mjDsH?
zyKcm2Ak(p1nfHf>dUx%~wdY!lP9B((qL`Pp{LGJAsh{lTq}{$N_DrE>@9de%qB`3j
zrt|jQJ$rBUV(asr&Ow(QH*g=Bp*!8_)x<tlRlAHux367TqcraTPa(rG7Q@RWIiEKa
z=(+q?3;f6x=WcTT|F+rLZ=8%1N=)1yUNwt1$lMXO)JpvOrdMvM>Ne9V7IUvub7*SZ
zoA+z}#>HQrrb=}!N;U3KYTsaTf9n?O4cdRA1r2OOt{;h-<0WAF_5xd4=fTe>8dZ4f
zle$Z<vd@oTpJ{SY@#_C0zqwDlTu`&nI?0|S-qE3*TzM$_zOeq^dw<@ie>uJ?VE6v5
zwn>-t?w$2?&AYgy&MfHKe6tgd6S+L+>#8i-7Lvo&aVh&s%pVr5Dep65{vTbUF<CWk
z%I8x{otGs?9=Z}V(e>Pi?!NFZW^!kL?=Vidanx~po5z&hA1iYo#OP~H?Ed8Y{Q5<y
zmhB1xId86d>-{_9ys&-lX&uXctGu13|K=A=km}o9rFzfnk>a!3k9(!7|HiIW;=R@!
z^>R<AW#G-T85cO#sD>ARm;ZR-*GJ(!J8yrpZSoJ8Qhi{;rvKOPC)~EkPh^k&D7ov6
zm(>Ho2I-yhB`HQ0jgO?CrT@7)$NHzm`im#c6`MYHv$$VM`)U*s{nbp%#roOGvsT&B
zm%l~c*SfLj#S6K;s#Rwiu3dV<7BBWAI<Dhmnd&K?ZBlRNSTKhLXcco-JWjlR?N6x9
zXQ53IPfy*pw(2T-W_0qsS6<?^fb&0ncs9<k=bF2)_q$|op?$f`sfA0|9QFG7x#{)M
zJ)Ne!+L4P6cD-Jer)gf9xW+{9D0|J5$?}oL5q#z!^g2HH+?#Om$d7U_$?QE9{fwS>
z=Gur)_!oZm^Si}b@&{y9Z?8(fI&0!>-H2y7iZ@e_?CZ(Yj>^7uy>(M+YvP=3aWC&Z
zQE{97XE{sylC`R$=@HLo9H{-<zx-h)V_2j5Y=wXKqnZA*FN@gjdNRa+_PN9vQ71TV
zeBSnN()|r{6FS#j?9p4;RWUJyf7Z>`H?1FQXC8lAF~_7x)9B*N#96t)FE6C@yEpsp
zZS_8WYR|ihmZ0FMRR)`WWSj4`)-YVayJ2Hco2KC8V-dG@sD>QlkiD8RH}^+h==Mgo
z8<X5(wzTfr9lcOrIa=J(rzi61fAMK<>G`XVH`r}{_4#~f^?$pN|3ByNEN$Hyvf-i|
zyU)3WDZRV*cWrsGVEH*yk^h-Hon&2=Ykqk8a=kWqzwTy04d?y8F3P`kH*4Iwx^K$e
zJ0EqWE*gHFv}xL%U#ow0$Qkf6>9w*>kajh{pl2Vixt%p^(tdj$H`^G8M47bYJDa0l
z9A#TAP<d}|#7qHop-%Oy$xrSEJ*cfWuDM(_!;AOfjH3Gu(XqRO*G*4TRle24Dks$*
z?a!x@!`70Ovggmd)F8X7Rg2c&IDI^7^45tRt?p8Ouf*AQIxJzG#(kt)%B$?`t>z08
zkN<mRBdZlFZu{?x*h$ID@+Ke7zfqi{FLUti=cP%T{q;O^Wy)?ye0XQSCy7yb)A}5<
z^ol84&h%Dm%#$;mY_8w(zf3mz)s&Ujx;hVBZdF>mKI)Ucq_WW8d8;+m*qiuvy?Gp-
z-?20C|9|f}2NqYpEBRi1<MEw4e{N`-P1^Re<FqT+VP(I$*55kW3tKkStVmQoE4=*M
z<%6E)yUp$#4>dXw=X60|;l(wkKb3d1CI#Q;yLC9M^H27Pd6Mb@ch3DcpK6e}NZ#YC
z80*SUCv#KOvcnc`mHM+}<B7Bzru+@}G+91on2SfWUAlHvob&$UBl#2dt8O{4Vak)P
z85>^N88E9xepo3NwBWkXUgyY=r8hsV2|eOz`exsUonD$3Rv)oY&;0eW?UT>hxdJ<T
zB4ocCTsCyNz3JAeBIc{MOeYj2Hh41%i`&l-2{@>bS8BS*!tun4ldXd3eG9lR&;PQo
zm*<eYk^P!ft-ZW=yC#~|cP?_x?i9Bad>?vs?S&(kJ&voJuPs#L*!5lSNA`6yg-|h}
zi*N29o$K`J$ZKA{_b&PO-xg}LFi*4M5er$%-)~*`*R22dS(#VAXHGQktdxH^<E_)F
z-LJ&`gKGBg6Ju)&3=_<M)~vIscFU}1Yt)r=r@hw*4~*YX_iuCfo7T!Hy^n5$o6TFU
z>ALqznZ-HfE9<vyFJe+pI@mug-{6~${Km76MQ;-x{GXv#>(26t+cRfw&cB*TZdaL(
zm@&+KR~lESoHu2oP4Kz7n%D1X#LDeC-7ff{Td*uc-NkFwv*-M3%lbGCfBub_X(Pg*
z_}*7CN4&(w@4^?B5B^D~IFn}cXgj4=E=lwIqFn1?^*rs=&b#M6&12<w6?0>|lZL_E
zDxHRekJCQLEwoNg-Lmsn_O^zN2R!rw(k!=R{Cp9bWwMNq{lLeZnBJwiDj5tP4k;_F
zJ{*3Y%ir`2$JM{{Cm+wy_^ev7^6!HBuWHG2HB?zIMO_m}`<k-&&GKTAqTa=2#xrNT
zP78dxqu!oxt=OCYZW~l?Z8h6|Yi+=3nNAn+^2<kFo-&fl`rfiJLh}m?mjs7WursgP
ztmgsygPf<=um9Rv*S%t|e2!RLc1~`G?$$L=*V^0@Qxly3tLJMWSAf7vl^B^5-p?ns
z?tQRw8uP7|7gd!-F$J4u-<mp~ZL8$=k01WOXr37McfH}c$&0ccX6QZp7<om3$3AM~
z36<_Owi4-STHMw*C(o0c{iFHUhK+G?Qx&Z^4@wjrW=tt9ui;$x`}HHO&+HOP>KjB{
z_dL#PY>>FsP_*!NYo1Q?NuS2eq8!JnS3TKtAj#0|R4h}X=$$pI7BBK>se0DbT<H5o
z;@P5eSI;y_pQ`QUI<F{kI*0vafM3KWxuWiyr;beD_^{%yL9mg_6wBW0UGAmV>Q4RK
zR=E7Lx$4!Z@aY0f%S2ul$*l{0zv<}Qmlvvii{?2km|CUI9#PKR@G4#Nk3!d(cN>#D
zC+<uX*qONfVA2dG2EI*q83S&8&aV4$h}(8ms+;2JOy0{^qAOigQeDMT?)=g^8Zt+}
zrn&#oLI!)q&JYjYoK3U0?e0suHDzmo^zIDLhMzX7e8#J*l|R)x6eavUKP#`$Nq^ND
z7s+5F5e7$P-}aqC-yTn0b)cbOcjViWji*1q-j;bYis#bFOS^YE?cFZc__O%6e9F4`
zvo~`ecrE+Ta>Tpm#R?IdaQ@r>V?>hPWk0vE(7C?u*unxnt-@!D0+&`V6Os;1+ibi#
z^61(d)8^jL(LO(a>!Sx%jY9v-*X&-{XHaqFj@<^&udjJ_#B22$REyo*9&z=<#uA6%
zh<n#~(>8Ap-Mu%ZS>z3y$%f;B+DoJ_o_S}Y+U?l6t|>2mqwPnR1#zM}^O_Z7JhiQv
zxl|;ZCa-O|GUL(GgO?8$&N}RB9r!qo`TdT^>1?%Y7nm)1;eCUnwONJBtZKu%-O&k}
zT(kbp$y>$EFD?~e`Qt<PBgV*p`1~rbKg+khHkO>S=w}y?-Ty^z%XABNnSJSYk&i!Z
z-D?@RKv(a}qYu};<{w)$OTPU76xQt8Y_l2B7j~-O{q}0!9G*iPl0xhh#Z-S^DTulk
zFZ;REKFBB0(<*<;Oo5(En{yX4rf$A^MC7G=R4LP9FQvq}#UCv-SCo{Bu9S$}pM3Db
z1$`6yQs;fsO?M}5PI|o~CWBqNGT=1_@BfW;b`kTI-Lif4LI3?n{$jQXPq{*lxHWE_
zyCZA%GVL{QOU}Gswd>lSgCz|AHmV%AO6NHE)2I8Io=tv4Vpo8X9IK;m<LvcwT&C~}
zuan$Ur|ZeRl0S6Mo@90PkXaVDIUc0_zWz(&!rWP#mUBkCNNy`XFri^dO3K73#nO|@
z>g+8uj;UUbZfVX6aV?ox=ir&fx@7y#<Ge0Mbsr|)y<l*Q;o_OMg?8(|9@<uW_sHpI
zkD8NG!(_H*TQ7-z)c0_PtB3LWYp>GZr<F0!toIeOGC3k8xPME_v~riE_jikr&0esK
zS2u>Oul<>Z%G4X86Mc^}ML&LZ`G<;L)0L+i^HO{sv7P!jJKS`->W(sVmuZi(oq0N1
zZXBw5VYy@RCAB|uqvvg%(=cDla<_b|Q_=brk3Sw>$MxYE$C{|CGech`tD9)W?P58x
zZ>9U?s=aetO1bQId^RYZDgNQX@zBrR=L&X;%Fhb+IS{5*`b~Y>IT=o;lmquBJyhMw
z#<Vcu*?*svo)zq~r!gIvKRL!P`&7E0gT@?<nLgQDEB7?s`;@$NC&$YJ)AY-@_@XzY
zmAb9Ynr5RP`)41I+Qyy`Yscf|>c(HU@%x%e{*#kxXp&mc<iNf?TcvrzUK=S!@#wd^
zr?daoE4uQI;{#*mx=`nM`KOglZd-V^2!!PNU#yP$_<P9)*N<<vEil_vI&bmYbK(*i
zC3W*Hii`xDPa0|~Ebw0Ap(iD=;6ZQi{IWCCeth`)XYQ-l@up`kF#2WPX%3Te|2zM6
zYq`Zd&yDu`K3THucea?q@4EY{P{%>rC0x(-q>ba}Te3=D%IUjpY@}MC&Y)0ixP3vU
z<cj|fqu(0|=z7&&D@{40&SUfHx7N*FH$-#Pymc4*=p-@quH=@tEt>T5uSARI9l=8Z
zMcvE(O1Gr@d?=2Yur7*I`0Syi3;YWsE#oKunI+2c@KpP(cx#U-lXS$R=PzP8m65jp
zak1AGTc<7G*?OW1ZU-DVXw7M8v*mql&+kizgT62|v_4kf^rMKGuY293bvv4Um3}=t
zrT=uBx)sxP-OdFLzgX%x?(w{Oa<yvF!qx3>Ke2vfH=p(E2v5BBX{Gi@VL|>2_zqem
zMa?|Rog~$G-}c(^xRslfj9c!6IeM&{ytVhZa!kmq3*RSQNj<+LXT`i$CpoF^Ed?*G
zmld2YO74oXJjfxitkgeY`VXB6UZ&-*y0b-kQ?h63y|}1#+)H!i<7?iNd#7<U8~k?r
z(2!{{>*CrDOGZY?u+Z--f*+(cNxpKJr{!R&@L#6eN>qK?n`^NF6PsJzIOArodg=A}
z>zBGkKWkp`w*9Q+J-3m8{nJJ})#7WbKYnX@{n>h_Y?fk4Sy`8n+8L8wBEN)J^UPhe
z{>ZG$-zRj&`1@SVm=nSiJ@M3oSb?V^3-3#2O?dqHY>#sBp(XawMF(ADCf&GHt$*su
zU9QQ|`xxs&R&N&GA$DEg_~K)i>lwz5_ucb@-ttSlTEfaRH-lg1p#J8n{~KavtzGc5
zujYz}!QX!V30Jf}CFJN!Bm`(a*nCp`uKcE!MYGm#$xUc>JhJlM2i;)T%i4c?Xa3}i
zchEFDm{hx{opS|KxKLU7@k?cEk99OxyDi<TXUe~|!9>e>acAdx`7Nng)BpbG*e}|Z
zz<B4qW!Q1)r_DNxCV1R(5Pq1R622^Y?(fo~U(EYwhH#}ssUKF(yAW4<{^cQ^g(0nH
z6fK%nv|e#4M6cO8C;uta*_;W7XHH5xWIeBa@pt(j9))Sg{^c!>+Vb+9&zbJ&y)BcD
zTs*y6s>Jb;s#}%n_stJG`z#mSO<9?uzU*;A^m8`Pt$OpAUi)AB<+r@DVYQMFQ`g74
zAJ)y2+UeUCaipF(#p~HcndIrmJB7IHBCJG=*F19C+bqemds*$>KlY5zKX7X-Ow8T5
z_0?42$lEn%mkSrmZ((rGGVGZ0bWg(T^=!<^r+04_;(GE?J79-AYyBMVWlA@t=l^?q
ztn%2`O?CkfwuY~HxO&@(HAgk2VpSs4gY2iCx|BQDbKjrC1)0ljQ%~BIOy6k0lowSJ
zZPVN+l=b1W7Uw-SM};Qt$-n2i>O8r2Y#)>A0oI7~?qN0B;sKX#)t%9QD;O|~oxe`{
z+wWbkA2{%;U$dB0nxZl3-HZE=4RS7eJ^ys)(}c>fdAl!ae=$=P-nG^4&6=a9cRTXl
zJyj$ye`~w`kEaKYPd8BFP^ep_adpAZz*PB+*C!^-u{N)H5h3Dm+r8@W=eNfajT~32
zMZNHv_2dPQ<?_Ae#)t3ye3+r7eLF?1S5_yBZ_D%2r&qO37VvCO-QTh(-zZ!C+Jqo;
zw+r{R#BZMa^6RN-EZ0xThwuON=q+A2qtEJzzq{)Gh@OWwdle2!91<~>-|CgKA@p;#
z=GJQqum4!Pmd~;HN`=dLE$IZyep4I4z`ZjCnDpK*nsWYsdCLlx#Db}ncdRzA<<ptN
zWP5+tmM5}q;jWgwJ}C@|S9GI0TG{3`uRO1G@<014;}1%IbQ~jIDm<GxPw=%@*pmP6
z7H2+*`^CJ@-({W1BrAvg>*laC+U{I)UgYGHudB>oJYANj68kLKQ*%lDHQrh4b9Ns%
z@b^z$-1@twSJez91Kqkm@hhe%)kmD}_pbc<Gg{E{FrUH0rKv}&PM)(d{^PQ-vT0Y=
z#Y?|Dyd_e5H1#f-|DKz@lPAi`BUv|gWup5}_t1-bru*A0T)rkya@iB+C0^YV?u(di
z`uk6+M&!&LIR(uf6P9`N*4)dTvj6KpzlS=#=?j^7RFZ9zb^1#l)UJ`fZBZ-yZU3pK
z%&S*uT%514!RPQ|7F~Z+Tf@D=$2o*$KWr1-_i1%gpyJOhf_9!CuU=<a@$cD|p5;Zk
zJhSHePUv#*UFe$7#Bsdi#iB)V^CYrAiPydFxZ)e$c5Cgq@=KE|TUpqGcg)ZHIy<?#
z*t+(*$!aUMo9s9EgH(DmvUUFMG38!)ebcq`8GRnYza==ul9OkM<b(@cpDma?-AVT5
z-9$f)uJ8#bu0B~{7@eV?`gr=<Mz0-gOC}w^$>*sR_|==gLAf=>;N+#)>igIKCR?QQ
z9qM?MyY{+)K(PgL*irM%rry2v7V`=VggocQH(Tkhl3uV;`&-n0oxb~lRkJd+wQoO-
znR?;xGU1OC81t*N+6=EAoNV%7rf^oulrK}1Lm%Gv@b`^ba?mj6a>mizODQ2Kw<o+4
zJO63<>i3gR&fNC&&y6p+$@ir*KRx`YWh8xoFJ##^Ce<IIAK(01>}nu;|1{H0^XrPS
zG5ggtOF!*9{_aYELDdEY&wDkZ@j~bKxorAyrT(}1wwB8lT(>{pvT-s>tKPWV&YPe0
z(v!n^IUWz(v?I+kIiJ<sN}lQXU+R_JmpM1SzDk<yyIUxFkE!hb=Wn0Q3b+^Wq~iJW
z?e`0~LX>LR)7C$h71o|rc%(D;&g1WVDGjOi-tocC>QmJ_lez<{?!D~0wElHZ&6Eiy
zfhP`XcPW?7$X#*&eA*3z-|BZ?NFEn(`*du|Mfp76l)yD#EOl*~Ud?pPTeEZ1Qpbhu
zJ7UA<ecXA*F>^CRNJU3pQ{%ga?GvMVZ&`8jTw$2(w3Bt>?6M6nKJv8puX=v|TDL>Q
z^6=XqPAXT3Fn#FxE|hQhFoki!Zo9dYa}FI)ae4OUZugvN7a#0c9rkcJo6+_RG4E^N
zE}Gd_DI8_W7qI+%H<pWQ{xOTGtt?-*?X4Em3$$6i`wXk^tGvxK4LLttoy#mXll$j9
z;Z%<W`g6|cZS8+yd;M_lsihBlXO)-iFgbpF%`M3@k}Fp>9v6<uOD;R7IB#Nm<N>q&
z;u2?F)27@{m0SPvoA_KU_QfkFM^1ZWdonGnJxK4R#omLr+*f#VJpQ>zdfx7NO=h+K
zqrE022^?EmxZ#sOL+35k_O07<tJMERSA<$AHzc;2C@<c;jPvb!FCNK|qjh{Tnikm+
zr*tdl9?7|$KV7`!|7oAJFsEyA!DrWh+-K)9?Rt;P%<r=5oC)U4d(R3jYcH(cb2T7b
zV1eF)2l79dCy2P&F1h&d@%?WXeT%0G9{)V~jMLP>neR3hnV!%OsB}#GSJJfJYZ=$3
z_Kqysbx!MQ%>+)q{iyoHX!fCZsb}_HUT^mGql~)zjQ6|bZ<f6=mz&K{ZTY2pdF0Pp
zo+*!Wqcw9HC+~W)YKl@sPF&^vOaGN;NY=TUe3qDP|K!(=rk_;@m9~826aIMe`Z_!Q
zkLTqx{{PauUB!9#VUy9{V#UtQ#z%M7>|AMg>BK$pZ#I!0>SwI9<g-janR<7B*_XH^
z>&W}p@qe}_^`3Ds+ccwd`F+mW4#&<rz5lu<$M)j=smdYBTMT#axYTtc?uDI$nB}_A
zJKI-1N|w2Q?2C`i)_%5%liA;%NY}_Ooe|tF`)^Brk`=Sj@n@_3YWe4`+tHHx_)Y2V
z<%a))=Nkq1&z}0}<iBU9Bb$0R%$}St+Mvy$^@BSq_5Y=wdxDpO=S`G;d(*Z}T4&|+
z$@<ol*YCVlmA_F|L)v!s&X3#DF0ab&nNk0IujogyV}=ty<vyC-9kOYnqwM_se^<`A
zpVw=;>D>0vqs)`<AIhB+@c3BAtg5Ib0n1FkrU!-0w|}409ku)E9yK4aHm>~=3u7#n
zpP7B1WARFLA9mNGsoG}0W|gc9-_T}eeyi!l(K)|5FRkmX^zB{ax^#7Un_8uZNb6nE
zO@fgzTJi;#Oq;tUnm#1Rw{O`UBpBmn`-UN-qTM;@)ys(c*K9r&R_-+OezWvPu;iQm
zOw+vcYd4+VHZ@UZjpJS3N&`)1?K3s<TKf829J8)o`&cxu)$zWM9pj6r>-r3A;+6bY
z?!9bbx~MvBdeJiny8<Q4^*2v`ciOc3QR~cy86kUQmo3;ir6e*gtlIzElRtU$c=})O
zTYE|7yw!z_vuorcPQ9#b-ZHBv$n<H;`%TTaZ>(~^Ct>XsHK*USVRCT#E5j9LY4@9r
z*H;SM)86&IXTBrfu0PJ@O($pbUXKjiR=(@)yd#fU?b{{nJL(RKhb{A*{;YcU*@V5d
zi=Cfo)SKK*sB+AGHFw&{M=R255{0^=enf?G{gRw38XkPD{{!oat)VT_M|XNwzPV}P
zpOmA%_Cb@3rcXWBtMk^kU8jDxSZnn}VW-KbGxPa==Dd9wo5#I$;`0CR7FFI6S+J;6
z?GmHk*4jBU`XdC}B;RwcJNd}fqfJiJBYe-Uf2Px3cE6Rk_d2{%si`38&5Z0Vo7l2)
z6s^<3bi~#jU%|On^SM~Uq>IAY#`@f)zkCZDuYd6?e`Dux<NkX0?bC9<TiDH=@i4IK
zetW|fj^(qw4*XlYd#_k?-b*>Y6@PLhozg>Ah;i%QKJZR$m-r3i<(0S3Gpagl`+I1o
zJNw6lGkv}XKmJ)<{o$X)zn8oB=NbQ&$y_a+vf9Mtq>qtw<eO(xw!JZVXtb){l+jDO
zXC|{*$gQ=LtDP&JSiV0Ubp3gNjveR5UyZ7vTswp&EDB(Lw{Lcc&n@k5sZ+$nI4dH8
z*qhdHXBB8vi(4w&Zhxk7!eQsqTYJ?^BO_y{?Xc?pXk^cN=xg=z<<UP|tb0Y(o>qQ#
zzBYBK!Tj$sJ(0eK3(j0<J-O_2&-z%et;@wt{{Kv@ciNOJzpc1`p~8V@_fjmsO}B{2
zO10~8KK5$+`KHBkPi?t+=Re|jw>9VgN7l-ou1LENJNmOLJGA;XZdNnwm9%>v$#-jC
zVT`@o<?03Bgp>vE^@d*9sgygZhdua!endI*#n<!re-y|>PVT-{yeA|wv%b1{vi-fJ
zt!q?Xox1nqoZHI#_YB@H=PZ>F+#K02zvbqR-z;`%Ti!fWJLGwC!j1QnCL8?<XBKHH
zY&)i_T`0BH_y>Q2sDVnQc7=H_i(OdYqeoTW>sCE~sUv#!UCN^N%j>^BRueaFWpEZ!
zy?pJKn5n(xI-_H1X~+8R+O{{C``S%&%lM(2Z#rk5eeS>R>tgrp&q}UxRA??ZJoU7-
z!sbi+H>OShyQHwa!f49r(yXRFkw8z|+21}cSYUj7mO}7jH{l$CIQ^~m+dOScpKZ~;
z9X+M!zO(PX%|7Ra^=3&w3SeGbnE2jy5r2=x;>DA$$-P%FxF31T^sP|Q$+|;}90j^R
zb3HR|(A|Ail}qz;&+)yoX+qh}g+)_;8ns=Ed~}n);N%9j%5seggIN^`lbSzHFx(;X
zG(I>ca{77MH4mLjIR7eMnt7>f<F1bTijBGK>1%w?cKcXdxN^Ko&1F63srU7q?j>vg
zuhW~DWiRx7iG)1=^VwG`_FXDoq_@DbJn2i4=(Z)oj}M5>P!yRI)$RK~@Xb^184Wpy
zR=50*`D8b1qNJ*@#ZjYWyAvh+{=AEmu80fc2zqt;g6*4qmp<*Damgv_gMPHw!LF~Z
zc^N-!f6Y^i-=X^Y{cnv%#S^;?e=jvQDM;M6;pIu!ng6S<D9&A2o5N^ssJO9c+9uas
z?C13tYh+9IMFs5G<Z>f@x`$xtYsLd}&2IZB_DeRoB+F;-i`6#mm2dlIaFW|ZC*a%e
zUmc!v{e|}*ntJcpv3U|z-Z@HJxV!Ie?Rz_2r)%1r-}B6ZpDZ!jchPOjuk)`|?_O)(
zex!0ik#P44qtbZ}uKBzC8S@r>`fFWz-g|lT7yomIkDL$Hj#!s+`t;1#neB_p`E%r#
zMQ{9gPHW}fzw@MyAKp_r|FQOsV&Sj9<PE%1BTwvAGk(8F-{VSsU-|Pt^|#9O)gu0S
z7mE6B-j-pjo6$1S^1At*k{foGAF>l3uG?|+VAE>#OJ6U!ZVK1_YP{EDix5Y;zCjlE
z-dmwgyUsPgYdBY#t1)>`rr?(!&W;sce*MQ>lBaQ}oSV;??%I9b>5G6u^5R4J>rbr{
zQM3ux+om2-s~uW-F`#PMohO^_?_{>)KM}$GdB+Z|mxAt!FE#{fpIYYpcmFc?P@R7p
z&gDiO&(qtpa(TFCiI~xe$JcL7UD+wM`@}1+W)6w4{Xgbr6ui2<VdCYr&)P%M@&r`o
zz0cXeZF%(0)CVp***sjt-oN5G|IDY>FCf9v=TZV=nZ=IHQu9nF9#~v%Y^4~hy{i40
zDYKpMq1g*EufEfNbbjxf!(0oJLZgqx-gxWzsYUb5uSee}PMk9%ZFY&nj+-*)7wLp4
z#m&m;uTh-YJLBtLF*h}}`lI@ix9q;!7z8_dL`-;nWrfF@H`gPA*-pN<mx=RxEX68h
zb>iF63L%XzPYiFGU$O9V^gPP@ebW_X#`{{)Au%7gmb^$_7_+Bgu3h$YAIH}|Q(oyB
zrt-9J>bmwWV)KH-?kW6xGg!MG9M~p&v-A1Ip!}@|GuTuwH2yLwd?=e~^i<rb`E<6o
zU0~Yn>V@{YKQ~Tr*tjx9Ra)T1E<U{$TU++;B5zLzUs@V%|2nqhOXDB^G8c}#{d^C^
zSdC8g>*f8|)qP{F9w+%}lCtjx(fbK`qCp(GfBwj%uX{0%_xO)rwI`R<JZrjG`6iHQ
zx#;?tN^@SxS~IF0N?-UaNH{1v;_<<)8}DpN-Tvmmo`s!9g}9Eh?3XN<;o)*=rNx5s
zIh`Gs=3D=hSK0CYtZUSbLLq*W-}fY(J5O;e@wvZAz;AJmXPd%iljUsduS@Odj`+Xn
z!;$2i;9V?66T~;`nc0OmUpjt;ZOMGeFpVBjox{JAL)lG!zTR+g7weaFLCX)z<NiJV
z%{xypX>PmVV>6riJEr_IuAQfGgCkr?Ol_U=qoVepm^iWEGoh2BLK^4&e`6DCvaoXP
z6rt%~<ZHijJhbU~eK5{=vvFO4n`Hfhmh(4Wd>78GUw5$UXzi!$Lz~ZYHu|p1+W&KF
zMYfanxyLSdnb*AxJF_oGQUAt|`PJn%&8jnBR>)rcYbe;G`y|{uQRx5cZMXk?@_lp1
zX@X@}>xSJO)3z5xEpb$-{eNt+-RfHVE9LL6WZpZKE+J$3t1?{o>CYKcqWg-irOcN}
zZR9-ZVETE&ys7HW2X7Ql`J*Hrb0zNMRh>mb;){09w+~hPa%dLwM(Y`SLNl9Lt{W|Q
zXmWyip79y`h68;&<)vEg{!@6fD0=seJ=cAA_=fP=zO<iP(S6hU`wljqE|-irlTCLI
zhL-)#Q||Z>wpzIN-u9bnZnqO>rfbiB>eui@jN_@S%BG%w4Z=%3tKOw7U#*-}$+GQ~
z*kt~Tr$2@$e|uyW{#|<C(_L3<z6hqz^6HSVlRc?@u&y~E=37MF?hRHFN2c}I-j}(*
zLRWHW`QyFk{@m8@U6DJh`pR7fCAHO?JC2wiZj?J(7cFtD^nBa*csY*QSL(zWFS)i(
zJtoC^j73gXAoXS6iFI3>8uq679d0<;D0sl=n&q?4M?1~ESFCqX`>*N0@WHwr3N=mh
z``4X6Y1Mdf%l~W%E=$e%y^a?xL++PPSiSnd+}xMZ_9ovdMMU&Ot{0t^sHmHHX&pyv
z?*x^%35T6B?z}K!UGO)Aw_1<+=E9Xcr+u>?J(KaSlD@87_tbTITzu~ybDNLxTy{q<
zvCdijpndt=qYFQr)ROLZ`&Dz5=ilnQ>h4{(X9@&oT#M<J*#5hyJY@EbZF8ov`8`^m
zCjYu}a_t<aPY+oBUYWh>@?Yh$tKrWIY$DgI9hXS#Te4|gk=Ail;fp^@4*5-6y}@-)
z``_lvucL0x-?b>pXmb{?hun>}6&sSe{L-X7%fe#jEYfrBTwuFLIr-t;`iaXrPdp5)
zoAhv}`s>6Gd!}9aqZ%q0oyYpGF!jfVYu~T#zx$&rOklEORqmcC>oWNJW~KNU-YQnP
z>il@ak_#uU2CGQSsNN@aIqZ6nm1AN1n)?$yc@C}(eE#f=?cp^d3pKZ>Nz3f=;{Nc1
zNh?xS&|o%?@-Klf-Kt}qk1MS0-Q(Y}DQ%37+4jimU_#|YW}A15Ozb`Wvq_{sK68Cq
zsBNLpw;6vf{Ic3OlX=gNos5qhI)1b?Z`VF?qq%0cLS^vw+SA3II^R1|&g}{Rar&E}
zkl7BG&FOQzL>U!*xtlAV|2*8h<A`q$)8TDOR@-=9T+H!VwD86D&TD=tyytt_gED1L
zYHjXrHs5e8@tx}Qy2k}OWw~M_1Ttm+e16;=?Q9dVH_7?crsglKViHMj{jaO~-^>fH
zsLKCxqxX@KO}{sP^=zlh&a4_Q_ufk=`|Cbyj&6Jff2zZGZv{hn)(5_xYC;L3vbzpi
zyf<_>_~Xt|mjxz@K2Oh?7_D!c_^dkSch>4--feGv^L@N8PKox33!8f8(neA56>{=x
zIq%HQpQs^qPm{AWZ_kZQsfiJrD?Mfw3M$Nf<X!burGCzrkCQhFxPRvlkkXo{{-t1B
z;Dk#j`}OUWYHkV7uxMCh$SlxN!H}2ca4zO&cTz<7ue-|+wW~1Zcdm_f73F(T{CwUM
zb^p8TtL3(=df{+q!~Xua*W$i0*)T{x5=#=7_MGf2uV|BE;4YW^Ur6fVx#p!0@|HTX
z78XA~_{C`9-&Jd-zFPD6+SVI(7u-LpFdeJ1yyU%2=kc-bEY4HSG9GdI>}#ACY*_te
z(vxfpCJU*!HU7uTd2<<0g$2+1q_}MYLuuIEzA(?VEB#6*s9xK0X8IEE<pSr;75bk$
zl<f_9lkiy3ooVxuw+)|s9-h#v{c>r4$?2;%Dz?_oo3v)i!N~%;Ztt{SHEoj(U1k*U
z^-S;gkR!Lx8?!#O{P%Wm=Zo!P^WSXBaIR5*{91aacGt`Nn`IIof`9DtX<S?HaOCE<
zJ14|GcdSqKZ<^5}5j;`#mdbC3u>D_q<_4V;>^$8jth`P_Q~TKa8TF55uKQ|m^Sa>t
z3)0=20y7(muKiyraBH4eIM0>r`<n7M^kSJOT)lhi`lL$-Jy_mt=$B%fti8E;miG$p
zycds{7AuH*Ot#!md8gR;K%&BV$5t7$qR_AHGtvu0N}Yb%F8%qGhbNQgsm}e~r48b@
z4c`eC2G+Z6wH2E0t*zy`?d+VS=Kqi19G=2qeLk6gztx=1J9=#M4+!1puKTjaAoR`N
ziIcd?x9`4nPv7FW=ZBv!1CErZ|1O)8?3k5Urg>{p>%rD(Y5aCV+b^B^5cQ;3eQJEe
zX@O#eQjPqL<@!x4{<9s=pW5fqvTx(@)vEsX{BwRz>8S6JKbDuc!c6qv=P4qo-2w09
zJcII7q<lJNrQZ7TYx0%e;A@vGqJA%N4&~R0I-ObE6JhPD=J&{KRp_ETS?(5z?%&ai
z89w#N`koOL|GB5-g+bYZ`OALAym?i}t|($~QhuhM{4TlJD$ZAvyF#CT^Ep!>YAdn)
z$xZn$TJtY<DbJ2#f4F7JiBt)}rT_OD+Wc<5e1`o|1s6ZpLJPT7BCU6tiw>EKguIsL
zJIQ^+LA#!*^_kMMC4#NBmFyGu*1DOpNAX686&?8Vu=wkz(AE38@--&eAKMgVwe0p$
zTYkfjv47b+G&e7C4Cp_T)O++?&L!D5<sC)skEj0qvh|axLWx__qQm@|IjP3U`Ffka
z6qVc;y!TI0;_MgB`HmmI^k3aEwf2w2q?|e}?P<DBXVUiFb12bxI;o%Qx%OMZ1y}Zn
zu*_Xpp#SRwm)o{(t0PP8xrzdGc{+7+7u27=>A?SKT6oNl17B|OUr$vm%vi>F<Fv!d
z!~P<{MSjQ2?Y{M%Id>)d#~WFO_b+m0|4#n&N-tH~-Sh6PUD>`CESx)MScS-aY$*A^
zVpGh@Q`?1GZ`!RWoW1z*Wa$vK?6Wyl=T7{~+$c8n;lJ==_MMFn7uT-%RMnWX)X0Q!
zY0-tX$M*_|d_SRj=5LsCv*mxj=FOX)s_wsFwPS@*$%l<?=TmKi%lv+9oLt`PyN9RY
z``To;iwlzX9cH;?kojfdw3Ij{`;w1kyH)tE-h9W^yi)e-O3@m{qZ@9|U92ZoB;L)w
zRpZ~EnBS2xJI|H>cdOdZD#8({=rlD{G0rr_KKt`hM+s)%z!|%D`yVP=$a$|;r&m29
zf+=U({h6LIvfSlgCN?B|b6o3JZ@gKO*?-2yOQ(K5`{cCwd%<bmwr@+$OP5_0nK8pF
z=YgDY=ea6_%2&1j3jXa_z9ep;TUyziULg*%b-ms4-xZdKb+39EHpzg2$3M3IqrkyW
z-}V~(*>Akk%E(+y`tQsK79nv<mQFT^lYXr$-fP0<d(AiO%OcUKx~HBw^xP?Z(A8h3
zvGn$|Q-5a7sqj8<BC^)_W6-^%TMH!baI9Q;`}YlnqRB@4W_50^aaxo*{qK{50R>+Z
z_rH$IuZ%2KD@%VV6}utva7w^=4vpWuG4o2C^d5J7HeA(VSQ1d#EpxQ%tICIM!LkqC
z{N|p|*!wVPMz=v5pX^JHeTU4&|0Se6vt;|UFW~IwE82Iyu_)<upN(B{H>5Q%D<{e%
zXN9fr?wEF)Phwv-D9>P9$uwd9nz<dON54I*&6pZ<{`B73r#+u$<>u8la=u#A?Hj|{
zv|8!f`>T3t@f`hG_oi}ZJ}}<re>LTD@p9h@7mMeyzb&;s5x?(s<YAWGV&?*P?J45P
zN$U7HVc%}n&l`^T+P$^f_-szC_P5CmQFV9TyQn@Y{dr=}QKeKqnVQ9M+fF9!cra(*
zrHrY~EXlGHqF$eu{{EolfN1};@Z1uSYl~-op7Q-#L+7S?7V)Vo_;wW@|2(lHJI~hR
zMDUvshL3wbojot5sHpybs>YA|vkWHoYWHaJFtRuNmd^e&gV}2LO>2p`BYbi$-a-xq
z@3u9j6qw3QJYSi!HCeXw^F<@Em4CCIEE0ISaZ$(~?Ve9L|0e3p6`7VQ)$Ebk-TULR
z75}ZR75S4FReQcFdg-60Cz_M|#9H9U(P`?fdl<hw;Fe2V@?(A6_mYe|28IRG{CaFE
z?-~_|@qX+q{e5e@<k9o$aZx+Id=p3(=6m42y7v_0X%o-}qwt(`fvC%m1sNE>Wi&J}
UGOX`gy=UW*^xsS%NtVbc0Jn2zga7~l
literal 0
HcmV?d00001
diff --git a/tests/fsck-tests/066-missing-root-orphan-item/test.sh b/tests/fsck-tests/066-missing-root-orphan-item/test.sh
new file mode 100755
index 000000000000..9db625714c1f
--- /dev/null
+++ b/tests/fsck-tests/066-missing-root-orphan-item/test.sh
@@ -0,0 +1,14 @@
+#!/bin/bash
+#
+# Verify that check can report missing orphan root itemm as an error
+
+source "$TEST_TOP/common" || exit
+
+check_prereq btrfs
+
+check_image() {
+ run_mustfail "missing root orphan item not reported as an error" \
+ "$TOP/btrfs" check "$1"
+}
+
+check_all_images
--
2.50.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly
2025-07-25 3:40 ` [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly Qu Wenruo
@ 2025-07-25 18:09 ` Boris Burkov
2025-07-26 0:43 ` Qu Wenruo
0 siblings, 1 reply; 7+ messages in thread
From: Boris Burkov @ 2025-07-25 18:09 UTC (permalink / raw)
To: Qu Wenruo; +Cc: linux-btrfs
On Fri, Jul 25, 2025 at 01:10:41PM +0930, Qu Wenruo wrote:
> [BUG]
> If we have a filesystem with a subvolume that has 0 refs but without an
> orphan item, btrfs check won't report any error on it.
>
> $ btrfs ins dump-tree -t root /dev/test/scratch1
> btrfs-progs v6.15
> root tree
> node 5242880 level 1 items 2 free space 119 generation 11 owner ROOT_TREE
> node 5242880 flags 0x1(WRITTEN) backref revision 1
> fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
> chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
> key (EXTENT_TREE ROOT_ITEM 0) block 5267456 gen 11
> key (ROOT_TREE_DIR DIR_ITEM 2378154706) block 5246976 gen 11
> leaf 5267456 items 6 free space 2339 generation 11 owner ROOT_TREE
> leaf 5267456 flags 0x1(WRITTEN) backref revision 1
> fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
> [...]
> leaf 5246976 items 6 free space 1613 generation 11 owner ROOT_TREE
> leaf 5246976 flags 0x1(WRITTEN) backref revision 1
> checksum stored 47620783
> checksum calced 47620783
> fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
> chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
> [...]
> item 4 key (256 ROOT_ITEM 0) itemoff 2202 itemsize 439
> generation 9 root_dirid 256 bytenr 5898240 byte_limit 0 bytes_used 581632
> last_snapshot 0 flags 0x1000000000000(none) refs 0 <<<
> drop_progress key (0 UNKNOWN.0 0) drop_level 0
> level 2 generation_v2 9
> item 5 key (DATA_RELOC_TREE ROOT_ITEM 0) itemoff 1763 itemsize 439
> generation 5 root_dirid 256 bytenr 5287936 byte_limit 0 bytes_used 4096
> last_snapshot 0 flags 0x0(none) refs 1
> drop_progress key (0 UNKNOWN.0 0) drop_level 0
> level 0 generation_v2 5
> ^^^ No orphan item for subvolume 256.
>
> Then "btrfs check" will not report anything wrong with it:
>
> Opening filesystem to check...
> Checking filesystem on /dev/test/scratch1
> UUID: ff32309e-4e90-4402-b1ef-0a1f9f28bfff
> [1/8] checking log skipped (none written)
> [2/8] checking root items
> [3/8] checking extents
> [4/8] checking free space tree
> [5/8] checking fs roots
> [6/8] checking only csums items (without verifying data)
> [7/8] checking root refs
> [8/8] checking quota groups skipped (not enabled on this FS)
> found 638976 bytes used, no error found <<<
> total csum bytes: 0
> total tree bytes: 638976
> total fs tree bytes: 589824
> total extent tree bytes: 16384
> btree space waste bytes: 289501
> file data blocks allocated: 0
> referenced 0
>
> [CAUSE]
> Although we have check_orphan_item() call inside check_root_refs(), it
> relies the root record to have its 'found_root_item' member set.
> Otherwise it will not report this as an error.
>
> But that 'found_root_item' is always set to 0, if the subvolume has zero
> ref, this means any subvolume with 0 refs will not have its orphan item
> checked.
>
> [FIX]
> Set root_record::found_root_item to 1 inside check_fs_root().
>
> As check_fs_root() is called after we found a root item, we should set
> root_record::found_root_item to indicate this fact, and allows proper
> orphan item check to be done.
>
> Signed-off-by: Qu Wenruo <wqu@suse.com>
> ---
> check/main.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/check/main.c b/check/main.c
> index b78eb59d0c50..1536c1bbbccd 100644
> --- a/check/main.c
> +++ b/check/main.c
> @@ -3728,8 +3728,7 @@ static int check_fs_root(struct btrfs_root *root,
> if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
> rec = get_root_rec(root_cache, root->root_key.objectid);
> BUG_ON(IS_ERR(rec));
> - if (btrfs_root_refs(root_item) > 0)
> - rec->found_root_item = 1;
> + rec->found_root_item = 1;
This change feels like an improvement to me, but it does seem to
implicitly reduce the fidelity of the root ref checking. Like if the
root refs was 0 but a ref existed, then we would suddenly miss that,
where before it was an error. I think it would be best to check that
refs is correct (i.e., store the found refs and add up the refs we see
and make sure they match), separate from the "found_root_item" logic.
> }
>
> memset(&root_node, 0, sizeof(root_node));
> --
> 2.50.0
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items
2025-07-25 3:40 [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Qu Wenruo
` (2 preceding siblings ...)
2025-07-25 3:40 ` [PATCH 3/3] btrfs-progs: fsck-tests: a new test case for missing root orphan item Qu Wenruo
@ 2025-07-25 18:10 ` Boris Burkov
3 siblings, 0 replies; 7+ messages in thread
From: Boris Burkov @ 2025-07-25 18:10 UTC (permalink / raw)
To: Qu Wenruo; +Cc: linux-btrfs
On Fri, Jul 25, 2025 at 01:10:40PM +0930, Qu Wenruo wrote:
> There is an internal bug report that some half-dropped subvolume makes
> balance to fail.
>
> It turns out that, the involved subvolume doesn't have an orphan item,
> this means the half-dropped subvolume is never going to be cleaned up.
>
> Then at balance time, a reloc tree is created for that half-dropped
> subvolume, and since balance doesn't expect to get a half-dropped
> subvolume, it doesn't check the drop_process_key and increased ref on
> already dropped nodes.
>
> The problem for progs is that, neither original mode nor lowmem detects
> this kind of problem.
>
> Original mode has a bad logic which prevents us from calling the proper
> orphan item check.
> Lowmem mode doesn't even take orphan item into consideration.
>
> This series will add the detection part for btrfs-check, for both
> original and lowmem mode, with a hand crafted image for test.
I left one idea on the check/original patch, but the series looks good
overall to me, thanks for the fix.
Reviewed-by: Boris Burkov <boris@bur.io>
>
> Qu Wenruo (3):
> btrfs-progs: check/original: detect missing orphan items correctly
> btrfs-progs: check/lowmem: detect missing orphan items correctly
> btrfs-progs: fsck-tests: a new test case for missing root orphan item
>
> check/main.c | 3 +--
> check/mode-lowmem.c | 11 +++++++++++
> .../066-missing-root-orphan-item/default.img.xz | Bin 0 -> 13468 bytes
> .../066-missing-root-orphan-item/test.sh | 14 ++++++++++++++
> 4 files changed, 26 insertions(+), 2 deletions(-)
> create mode 100644 tests/fsck-tests/066-missing-root-orphan-item/default.img.xz
> create mode 100755 tests/fsck-tests/066-missing-root-orphan-item/test.sh
>
> --
> 2.50.0
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly
2025-07-25 18:09 ` Boris Burkov
@ 2025-07-26 0:43 ` Qu Wenruo
0 siblings, 0 replies; 7+ messages in thread
From: Qu Wenruo @ 2025-07-26 0:43 UTC (permalink / raw)
To: Boris Burkov; +Cc: linux-btrfs
在 2025/7/26 03:39, Boris Burkov 写道:
> On Fri, Jul 25, 2025 at 01:10:41PM +0930, Qu Wenruo wrote:
>> [BUG]
>> If we have a filesystem with a subvolume that has 0 refs but without an
>> orphan item, btrfs check won't report any error on it.
>>
>> $ btrfs ins dump-tree -t root /dev/test/scratch1
>> btrfs-progs v6.15
>> root tree
>> node 5242880 level 1 items 2 free space 119 generation 11 owner ROOT_TREE
>> node 5242880 flags 0x1(WRITTEN) backref revision 1
>> fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
>> chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
>> key (EXTENT_TREE ROOT_ITEM 0) block 5267456 gen 11
>> key (ROOT_TREE_DIR DIR_ITEM 2378154706) block 5246976 gen 11
>> leaf 5267456 items 6 free space 2339 generation 11 owner ROOT_TREE
>> leaf 5267456 flags 0x1(WRITTEN) backref revision 1
>> fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
>> [...]
>> leaf 5246976 items 6 free space 1613 generation 11 owner ROOT_TREE
>> leaf 5246976 flags 0x1(WRITTEN) backref revision 1
>> checksum stored 47620783
>> checksum calced 47620783
>> fs uuid ff32309e-4e90-4402-b1ef-0a1f9f28bfff
>> chunk uuid 6c373b6d-c866-4c8c-81fa-608bf5ef25e3
>> [...]
>> item 4 key (256 ROOT_ITEM 0) itemoff 2202 itemsize 439
>> generation 9 root_dirid 256 bytenr 5898240 byte_limit 0 bytes_used 581632
>> last_snapshot 0 flags 0x1000000000000(none) refs 0 <<<
>> drop_progress key (0 UNKNOWN.0 0) drop_level 0
>> level 2 generation_v2 9
>> item 5 key (DATA_RELOC_TREE ROOT_ITEM 0) itemoff 1763 itemsize 439
>> generation 5 root_dirid 256 bytenr 5287936 byte_limit 0 bytes_used 4096
>> last_snapshot 0 flags 0x0(none) refs 1
>> drop_progress key (0 UNKNOWN.0 0) drop_level 0
>> level 0 generation_v2 5
>> ^^^ No orphan item for subvolume 256.
>>
>> Then "btrfs check" will not report anything wrong with it:
>>
>> Opening filesystem to check...
>> Checking filesystem on /dev/test/scratch1
>> UUID: ff32309e-4e90-4402-b1ef-0a1f9f28bfff
>> [1/8] checking log skipped (none written)
>> [2/8] checking root items
>> [3/8] checking extents
>> [4/8] checking free space tree
>> [5/8] checking fs roots
>> [6/8] checking only csums items (without verifying data)
>> [7/8] checking root refs
>> [8/8] checking quota groups skipped (not enabled on this FS)
>> found 638976 bytes used, no error found <<<
>> total csum bytes: 0
>> total tree bytes: 638976
>> total fs tree bytes: 589824
>> total extent tree bytes: 16384
>> btree space waste bytes: 289501
>> file data blocks allocated: 0
>> referenced 0
>>
>> [CAUSE]
>> Although we have check_orphan_item() call inside check_root_refs(), it
>> relies the root record to have its 'found_root_item' member set.
>> Otherwise it will not report this as an error.
>>
>> But that 'found_root_item' is always set to 0, if the subvolume has zero
>> ref, this means any subvolume with 0 refs will not have its orphan item
>> checked.
>>
>> [FIX]
>> Set root_record::found_root_item to 1 inside check_fs_root().
>>
>> As check_fs_root() is called after we found a root item, we should set
>> root_record::found_root_item to indicate this fact, and allows proper
>> orphan item check to be done.
>>
>> Signed-off-by: Qu Wenruo <wqu@suse.com>
>> ---
>> check/main.c | 3 +--
>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>
>> diff --git a/check/main.c b/check/main.c
>> index b78eb59d0c50..1536c1bbbccd 100644
>> --- a/check/main.c
>> +++ b/check/main.c
>> @@ -3728,8 +3728,7 @@ static int check_fs_root(struct btrfs_root *root,
>> if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
>> rec = get_root_rec(root_cache, root->root_key.objectid);
>> BUG_ON(IS_ERR(rec));
>> - if (btrfs_root_refs(root_item) > 0)
>> - rec->found_root_item = 1;
>> + rec->found_root_item = 1;
>
> This change feels like an improvement to me, but it does seem to
> implicitly reduce the fidelity of the root ref checking. Like if the
> root refs was 0 but a ref existed, then we would suddenly miss that,
> where before it was an error. I think it would be best to check that
> refs is correct (i.e., store the found refs and add up the refs we see
> and make sure they match), separate from the "found_root_item" logic.
That sounds great.
We indeed need to record the actually found root refs (the current
found_ref) and the expected root refs in root item.
This looks easy enough, and I'll add that first.
Thanks,
Qu
>
>> }
>>
>> memset(&root_node, 0, sizeof(root_node));
>> --
>> 2.50.0
>>
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2025-07-26 0:44 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-07-25 3:40 [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Qu Wenruo
2025-07-25 3:40 ` [PATCH 1/3] btrfs-progs: check/original: detect missing orphan items correctly Qu Wenruo
2025-07-25 18:09 ` Boris Burkov
2025-07-26 0:43 ` Qu Wenruo
2025-07-25 3:40 ` [PATCH 2/3] btrfs-progs: check/lowmem: " Qu Wenruo
2025-07-25 3:40 ` [PATCH 3/3] btrfs-progs: fsck-tests: a new test case for missing root orphan item Qu Wenruo
2025-07-25 18:10 ` [PATCH 0/3] btrfs-progs: check: add detection for missing root orphan items Boris Burkov
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.