"value":"In the Linux kernel, the following vulnerability has been resolved:\n\nnilfs2: fix failure to detect DAT corruption in btree and direct mappings\n\nPatch series \"nilfs2: fix kernel bug at submit_bh_wbc()\".\n\nThis resolves a kernel BUG reported by syzbot. Since there are two\nflaws involved, I've made each one a separate patch.\n\nThe first patch alone resolves the syzbot-reported bug, but I think\nboth fixes should be sent to stable, so I've tagged them as such.\n\n\nThis patch (of 2):\n\nSyzbot has reported a kernel bug in submit_bh_wbc() when writing file data\nto a nilfs2 file system whose metadata is corrupted.\n\nThere are two flaws involved in this issue.\n\nThe first flaw is that when nilfs_get_block() locates a data block using\nbtree or direct mapping, if the disk address translation routine\nnilfs_dat_translate() fails with internal code -ENOENT due to DAT metadata\ncorruption, it can be passed back to nilfs_get_block(). This causes\nnilfs_get_block() to misidentify an existing block as non-existent,\ncausing both data block lookup and insertion to fail inconsistently.\n\nThe second flaw is that nilfs_get_block() returns a successful status in\nthis inconsistent state. This causes the caller __block_write_begin_int()\nor others to request a read even though the buffer is not mapped,\nresulting in a BUG_ON check for the BH_Mapped flag in submit_bh_wbc()\nfailing.\n\nThis fixes the first issue by changing the return value to code -EINVAL\nwhen a conversion using DAT fails with code -ENOENT, avoiding the\nconflicting condition that leads to the kernel bug described above. Here,\ncode -EINVAL indicates that metadata corruption was detected during the\nblock lookup, which will be properly handled as a file system error and\nconverted to -EIO when passing through the nilfs2 bmap layer."
