diff mbox

[2/2,v3] Btrfs-progs: add fsck test for filesystem with shared prealloc extents

Message ID 20180314201118.12451-1-fdmanana@kernel.org (mailing list archive)
State New, archived
Headers show

Commit Message

Filipe Manana March 14, 2018, 8:11 p.m. UTC 
From: Filipe Manana <fdmanana@suse.com>

Verify that a filesystem check operation (fsck) does not report the
following scenario as an error:

An extent is shared between two inodes, as a result of clone/reflink
operation, and for one of the inodes, lets call it inode A, the extent is
referenced through a file extent item as a prealloc extent, while for the
other inode, call it inode B, the extent is referenced through a regular
file extent item, that is, it was written to. The goal of this test is to
make sure a filesystem check operation will not report "odd csum items"
errors for the prealloc extent at inode A, because this scenario is valid
since the extent was written through inode B and therefore it is expected
to have checksum items in the filesystem's checksum btree for that shared
extent.

Such scenario can be created with the following steps for example:

 mkfs.btrfs -f /dev/sdb
 mount /dev/sdb /mnt

 touch /mnt/foo
 xfs_io -c "falloc 0 256K" /mnt/foo
 sync

 xfs_io -c "pwrite -S 0xab 0 256K" /mnt/foo
 touch /mnt/bar
 xfs_io -c "reflink /mnt/foo 0 0 256K" /mnt/bar
 xfs_io -c "fsync" /mnt/bar

 <power fail>
 mount /dev/sdb /mnt
 umount /mnt

This scenario is fixed by the following patch for the filesystem checker:

 "Btrfs-progs: check, fix false error reports for shared prealloc extents"

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
---

V3: No changes.
V2: No changes, only added Qu's reviewed-by tag.

 .../reflinked-prealloc-extents.img.xz              | Bin 0 -> 3244 bytes
 .../030-reflinked-prealloc-extents/test.sh         |  42 +++++++++++++++++++++
 2 files changed, 42 insertions(+)
 create mode 100644 tests/fsck-tests/030-reflinked-prealloc-extents/reflinked-prealloc-extents.img.xz
 create mode 100755 tests/fsck-tests/030-reflinked-prealloc-extents/test.sh

Comments

David Sterba March 19, 2018, 12:47 p.m. UTC | #1 
On Wed, Mar 14, 2018 at 08:11:18PM +0000, fdmanana@kernel.org wrote:
> From: Filipe Manana <fdmanana@suse.com>
> 
> Verify that a filesystem check operation (fsck) does not report the
> following scenario as an error:
> 
> An extent is shared between two inodes, as a result of clone/reflink
> operation, and for one of the inodes, lets call it inode A, the extent is
> referenced through a file extent item as a prealloc extent, while for the
> other inode, call it inode B, the extent is referenced through a regular
> file extent item, that is, it was written to. The goal of this test is to
> make sure a filesystem check operation will not report "odd csum items"
> errors for the prealloc extent at inode A, because this scenario is valid
> since the extent was written through inode B and therefore it is expected
> to have checksum items in the filesystem's checksum btree for that shared
> extent.
> 
> Such scenario can be created with the following steps for example:
> 
>  mkfs.btrfs -f /dev/sdb
>  mount /dev/sdb /mnt
> 
>  touch /mnt/foo
>  xfs_io -c "falloc 0 256K" /mnt/foo
>  sync
> 
>  xfs_io -c "pwrite -S 0xab 0 256K" /mnt/foo
>  touch /mnt/bar
>  xfs_io -c "reflink /mnt/foo 0 0 256K" /mnt/bar
>  xfs_io -c "fsync" /mnt/bar
> 
>  <power fail>
>  mount /dev/sdb /mnt
>  umount /mnt
> 
> This scenario is fixed by the following patch for the filesystem checker:
> 
>  "Btrfs-progs: check, fix false error reports for shared prealloc extents"
> 
> Signed-off-by: Filipe Manana <fdmanana@suse.com>
> Reviewed-by: Qu Wenruo <wqu@suse.com>

Applied, thanks.
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/tests/fsck-tests/030-reflinked-prealloc-extents/reflinked-prealloc-extents.img.xz b/tests/fsck-tests/030-reflinked-prealloc-extents/reflinked-prealloc-extents.img.xz
new file mode 100644
index 0000000000000000000000000000000000000000..8adf0071328806fa6981f6ef225084e517d1bc3e
GIT binary patch
literal 3244
zcmV;d3{&&{H+ooF000E$*0e?f03iV!0000G&sfanlm85CT>wRyj;C3^v%$$4d1wo3
zjjaF1$3n~+-*XiD+@YGxDtdHe$qq4$Wo>7O_CGnuIn8OLnT=x4IGVZ-w??mVmRMZ7
z5Ay+V;mHSAoDkyM-tpo^bS;x+v}A`sNQY^&CEysS!9&~{eDz@XZKXO?8OF)IjGm%(
zDUT1JIsHjj>>}y5qdSJT79-V@5Qr*LP$DwA@XNl^>?ShU!y!KD38l^=LE*nzF2}qs
zeB4zmhwleI8r9LN(p#cQeLEqdB4_Wgl7Qx-M>xaL`0TU`9k;qs6+<m^MS6#aJJLSd
zXM<OzIRBi~FXfyl0F*QQU?q@ef_no!z!h5sKrpOaNUmVT0P^oXmh)$zOPv*4QhZ8l
zbzW6TbADEeo&(O%KRo~J4wOx<8^^<e!ESp;y1q0bpp0Zp`Ezji8TxNVYzELV*W8?U
zR|%W+)vd-|TFoh=V(aRn<Yaty2+be42Q*DQdpPQ<mUvI!(VdtnFjiVA<q0bk(%vS7
z`~+>KP*H>U4>=}6F*5_MhCaIzBlfo8uAWSd<n_9_;8oShBE#$g$F)f-{y4nKZrN;)
zrrqRuQ^codZ2Kn&sS)_gA^AxzCmLAKd9_IOcc1Y7AMt3b+k=e#a($+4(Opd}`U9d)
z^)tHw_!Y|aTnu2XF-^Q*e_bho&VbZZ?o^i0#1X$TDH6iIUh^XdYVJ!ilgmw-O5El(
zumD!#UK1_PXUm!foq5b^t%|M*tpL#EnH2)Cf4b*tRqW0nKb?DS;yRc6RF!u$hEc{#
zuV74H@cG>+hP%WEflaqMvG9Yi+k#=?=X1;)RqEp2fES1y0!CcOxHBNudB4H&Ibwd~
zg^Q=pEd>6f!;l~S^outD;V9&&z4zdU@RLa{&{9spkk|mHXS%NxkxrGeraxy^{Cx+(
z$hvBW&sc#`Zhts1td;TLkX1RsNg`CWT*I)*I@mk(hGjz?_yyPd-M$ua7B`xni7MSs
zWyl`7wxZ%$wBM$&J&KIc2)P<1(dhx?%fv+}{l<|u?L$$QGeT(rUdBI{<sq_+NoIb*
zXCFfHo;iS}I4TvO*KUA$IhfL*bB&b+2#;vpv_RqGMjsg7rfNsW-cO)wKmbe(r<on+
zHVI}`RRHmEs|Cy<n*v#f7E|LlnD>pRI`}U)a7Fj|kXDbn-0nez<tl-OS5Nd_LEnD!
zrIjIt8QWvqSzFu+;8ojJrn=f+bW{Ku2_X;=10zSJ(I{2dyLYDz1lqk%q@f`ZV}||G
zof&_x9#}S`A6KopUV|$79oH6MIL;iQZ1@z<Z3ipwXr*mbIE<^5pjL8})6%h~ufNC^
z{ZoXk;V`h%E`^1ttXPm@?Y?qs4N$(iN8v_(!97^1j5xz}LAP1$2$d8MlE6noMl>Re
zyo)Z$_6XM$t$LFEKukadE#&l-4l*)^!)pQ&8mYw_kMjNgxd*nQI${ow>pAWyg-&V7
zeY~@@vvyi{r6BFgsls*62dIJ6pHyrL+V9bwi^Y3AmVbJ-rj~|0%3qm%C-loCfyP>s
z$_FFfC0=|&kxq!1xk(2zP?u6Z!@qsjy-cb?f8VOdA*CL8lOP3H25+rghFiC_`b!X$
z&Yl>P+>~7sXasAE_)+#HFn-4WN^kg_vvy*M&~#=`LtcOqgLlop`KKZ3Jgg1KVc>0+
z<&7l#_?5OdU<UBg_`ChQ^+9bTjZihW`tO`4u={%no{n#yjOqf``Y4)h06-ANQS)Gh
zdCCSN6Z0E*)={hbEF|{|y>bH9q)rI1J($!(YD81|JB{3iNd4fYKY~4PtyP7Z3#>vE
zLbdE%`v*0S%thI+abigpJ5|RLJpSeZubQp{B{+2vJAP~H{7!6?%0k~ao{0;mLf$)D
z#8FMC2d0|gj@k<#OyA&Smcaa9Ex7m%!|kb9cc7H-YKcGH#V8vGFVvS$6Fd)cwk&yj
ze|9UrRDX1SMKK2wwAIrYlX^SQ^Gd1TmT8TquvY&Ww5aR#QgwM?S~ZIw4!K(Esm|Uq
z`*nSS2CYP0d>BeDe@8%d7RbS$9;pJuD7(EMB+vmZ%gK70s>Oyw9^pFXUsEVku?z;W
zEAMEG2zVPq|IQkY!446b#zBa;-T;gc5rz(FDUjCpB9(^S_qi=4Lhcn8?1nXaN5b_i
z6}`_lNEJMtzqR%|caxD_PZ1dBV+a)jInRU=V%d);03w3e2LWwuKn=GMU<zaf*sCE~
zw8jSM{|g%-wx@RYhOmjs?7iItq`QM31q?6of<epdmdS254NMF-Y&36yMN_;E5s|@%
zfH#0CgQ_{xo|M<@;ypweWJFNtZ`$UT1#LK;+BT|j?Y>>w=2+NG32Rn!Y<jomHMzWM
zEf>aXN2+<!Yz(j(W?*X)mU=1h7`^qlua_{+|D|=l(Eq-k9hf~v#Cy;iGC$T>v~(kl
zo^@2=_Kn$=vP0Dp_k^q>8Q{e{(7kB%9_mJ|<39=H&0Li1=wcrd{hN7URQks)E8JY=
zgM5#Nt%rY;GY*s%)x8WSes6E8vtAxZ0RAS&oR+(V%}@kdj~<PlKVab0g;!yqOSYF!
z)SjQxIx6Ic3J}%|Bf1^W`chZ0UCAHQJXyUHA`eXx3=I?%oX!XP!Wu*$$;>mJ5{x#s
zou?Fb;e7HrfU;p4?0sY877;Z1H?}-Ygcc31-G9Y(;OpjA_g*cpG4-G%Xr76TlW!S!
zD_4#Gi<Wz-DAt^f$j?hp8ME4^p_@z}4X#GX2h`Qs%L?IFXLWewf;y6DZO?DQx;R)D
zqU2WjKif$!=o6m%p`?{weM8d#!)lk-0@$>J9d6}O$qGTb3Ue-WLHSzqdibPv^Qgkk
zF!gSYFP0o14elE{;)bNl<F*B{@c>2Z@l^s&kpk_=eh+ez))ZImHa$*Ti;e$`XkLqg
zSP(U&{f0R?H;DC~3y7mGE!{R^vpI=_uSE%tjI~g#ih?Ln6gz3n`*V}gKOyr)vUi}w
z8AcDrm${MstR=4MxubM%QS6^8F9MtEA!AVMExun)-S{WhkF+ed-#7js)&$Fm9l>Et
z>G~Km`H;$$RF;JKpJ_{*56=Se53(~Yv`{#DSO;w=e6G4}ivFf?PPm}qFweC&J>2lA
zz)zLztUzz8+xjs4vKTGtziG?4_!{L8QLiD59#aYO9*$)ZFa;jOO2JwKYkd-kOVQI;
zM{e<1!HQg4+0Hg@>Xqpf?S057lpn_3=Q*31|3jqcYeRZ6oCA*rIHb4n9bHL=aZS?v
zvrOjS^s&Pojj4!3PjKdLh8=5AL<g6l?^7{n2&hUjj}ijvA<*7a9cJ((wuSo%6p{8b
z2RUFS|5XoAZ^iZf7jooSYPp3n={o%go~S(K3LWq=i%=CTU+CY04N6B^ysy@Y1#jSH
ziU{ZW*?VF1bbDRG!Y#kB>)}GE_8FUOoL)ydGOEw#8T^B}n)nXror4+_n)>Z2*|b&r
zQ4u*Sk<0g^s<hgs%!wKRAp9dwGen0ufCspP@ViVRqp@PRH@ce6LE~>2t7tye&Xdg8
zG2HN`QJka}O}{dIqcp84O#rE32dAf5zHVN^@n8B&66QNP!JVEEuM%12AX~oRPIcGs
zV=ps`>|WP1cQu?kUL<77wB&hGzYt`M0=lV#So{@_Iygr{o5Fn9<QgE@N(urC{1K(O
zsuhG@9`|vfrB??t>|9HMk+a=FmqKk8_QB99ArRuR7>#!y%L9G!{j6!ncOTk_*3awa
zkN;`~%$XI@o0Ahz7<PR=mH@5kl5ngzKLK+-m3B$(hLi;q`LOTgX`wh6F`)*>)ZNo_
z2c`k7Oh;fsno7uR0!pBEg5BgjohMu8?^T<uxsEH9WQG7EnG{3?3YSnGb%#uxvS*3R
zOcN1c_k+ZeD-zmXY~-g~9nLe(cf?$O-!1xD4^2f|pF=%NWj!N&xU+#xYVPr4=Z8T{
zu7*SsVq6K#K<LXYuwk6nWWj~%Xc?1)39VDOYaKOw=N0=&!mgmbv}Is>N11DabusTj
ze)_pSN^Ba6Wp^2aiQUvs2r>?R1ocR&>Q=DkRQbI>pgveJcmHt~_=uIcP3k7)2^GDc
zN537UWwtmza_!k~6snP|;niI(9;M^(9c^6$;QOu}{65*q*%Fd|N&ck?tRCcNBmAhX
z2F3n;IubrupC)EJ-_;I9Z+}0XRSRX-ats!dr@lFlB;jzLrz2nNCO!9`9{vD7cva-#
e`Wb})0f-rZs2TvH_gU7l#Ao{g000001X)^Ms5!_0

literal 0
HcmV?d00001

diff --git a/tests/fsck-tests/030-reflinked-prealloc-extents/test.sh b/tests/fsck-tests/030-reflinked-prealloc-extents/test.sh
new file mode 100755
index 00000000..63f692bc
--- /dev/null
+++ b/tests/fsck-tests/030-reflinked-prealloc-extents/test.sh
@@ -0,0 +1,42 @@ 
+#!/bin/bash
+#
+# Verify that a filesystem check operation (fsck) does not report the following
+# scenario as an error:
+#
+# An extent is shared between two inodes, as a result of clone/reflink operation,
+# and for one of the inodes, lets call it inode A, the extent is referenced
+# through a file extent item as a prealloc extent, while for the other inode,
+# call it inode B, the extent is referenced through a regular file extent item,
+# that is, it was written to. The goal of this test is to make sure a filesystem
+# check operation will not report "odd csum items" errors for the prealloc
+# extent at inode A, because this scenario is valid since the extent was written
+# through inode B and therefore it is expected to have checksum items in the
+# filesystem's checksum btree for that shared extent.
+#
+# Such scenario can be created with the following steps for example:
+#
+# mkfs.btrfs -f /dev/sdb
+# mount /dev/sdb /mnt
+#
+# touch /mnt/foo
+# xfs_io -c "falloc 0 256K" /mnt/foo
+# sync
+#
+# xfs_io -c "pwrite -S 0xab 0 256K" /mnt/foo
+# touch /mnt/bar
+# xfs_io -c "reflink /mnt/foo 0 0 256K" /mnt/bar
+# xfs_io -c "fsync" /mnt/bar
+#
+# <power fail>
+# mount /dev/sdb /mnt
+# umount /mnt
+
+source "$TEST_TOP/common"
+
+check_prereq btrfs
+
+check_image() {
+	run_check "$TOP/btrfs" check "$1"
+}
+
+check_all_images