CVE-2022-50434

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> blk-mq: fix possible memleak when register &amp;#39;hctx&amp;#39; failed<br /> <br /> There&amp;#39;s issue as follows when do fault injection test:<br /> unreferenced object 0xffff888132a9f400 (size 512):<br /> comm "insmod", pid 308021, jiffies 4324277909 (age 509.733s)<br /> hex dump (first 32 bytes):<br /> 00 00 00 00 00 00 00 00 08 f4 a9 32 81 88 ff ff ...........2....<br /> 08 f4 a9 32 81 88 ff ff 00 00 00 00 00 00 00 00 ...2............<br /> backtrace:<br /> [] kmalloc_node_trace+0x22/0xa0<br /> [] blk_mq_alloc_and_init_hctx+0x3f1/0x7e0<br /> [] blk_mq_realloc_hw_ctxs+0x1e6/0x230<br /> [] blk_mq_init_allocated_queue+0x27e/0x910<br /> [] __blk_mq_alloc_disk+0x67/0xf0<br /> [] 0xffffffffa2ad310f<br /> [] 0xffffffffa2af824a<br /> [] do_one_initcall+0x87/0x2a0<br /> [] do_init_module+0xdf/0x320<br /> [] load_module+0x3006/0x3390<br /> [] __do_sys_finit_module+0x113/0x1b0<br /> [] do_syscall_64+0x35/0x80<br /> [] entry_SYSCALL_64_after_hwframe+0x46/0xb0<br /> <br /> Fault injection context as follows:<br /> kobject_add<br /> blk_mq_register_hctx<br /> blk_mq_sysfs_register<br /> blk_register_queue<br /> device_add_disk<br /> null_add_dev.part.0 [null_blk]<br /> <br /> As &amp;#39;blk_mq_register_hctx&amp;#39; may already add some objects when failed halfway,<br /> but there isn&amp;#39;t do fallback, caller don&amp;#39;t know which objects add failed.<br /> To solve above issue just do fallback when add objects failed halfway in<br /> &amp;#39;blk_mq_register_hctx&amp;#39;.