In the Linux kernel, the following vulnerability has been resolved:
mm: use memallocnofssave() in pagecachera_order()
See commit f2c817bed58d ("mm: use memallocnofssave in readahead path"), ensure that pagecachera_order() do not attempt to reclaim file-backed pages too, or it leads to a deadlock, found issue when test ext4 large folio.
INFO: task DataXceiver for:7494 blocked for more than 120 seconds. "echo 0 > /proc/sys/kernel/hungtasktimeoutsecs" disables this message. task:DataXceiver for state:D stack:0 pid:7494 ppid:1 flags:0x00000200 Call trace: _switchto+0x14c/0x240 _schedule+0x82c/0xdd0 schedule+0x58/0xf0 ioschedule+0x24/0xa0 _foliolock+0x130/0x300 migratepagesbatch+0x378/0x918 migratepages+0x350/0x700 compactzone+0x63c/0xb38 compactzoneorder+0xc0/0x118 trytocompactpages+0xb0/0x280 _allocpagesdirectcompact+0x98/0x248 _allocpages+0x510/0x1110 allocpages+0x9c/0x130 folioalloc+0x20/0x78 filemapallocfolio+0x8c/0x1b0 pagecacheraorder+0x174/0x308 ondemandreadahead+0x1c8/0x2b8 pagecacheasyncra+0x68/0xb8 filemapreadahead.isra.0+0x64/0xa8 filemapgetpages+0x3fc/0x5b0 filemapspliceread+0xf4/0x280 ext4filespliceread+0x2c/0x48 [ext4] vfsspliceread.part.0+0xa8/0x118 splicedirecttoactor+0xbc/0x288 dosplicedirect+0x9c/0x108 dosendfile+0x328/0x468 _arm64syssendfile64+0x8c/0x148 invokesyscall+0x4c/0x118 el0svccommon.constprop.0+0xc8/0xf0 doel0svc+0x24/0x38 el0svc+0x4c/0x1f8 el0t64synchandler+0xc0/0xc8 el0t64_sync+0x188/0x190