In the Linux kernel, the following vulnerability has been resolved: nvme-pci: fix a NULL pointer dereference in nvme_alloc_admin_tags In nvme_alloc_admin_tags, the admin_q can be set to an error (typically -ENOMEM) if the blk_mq_init_queue call fails to set up the queue, which is checked immediately after the call. However, when we return the error message up the stack, to nvme_reset_work the error takes us to nvme_remove_dead_ctrl() nvme_dev_disable() nvme_suspend_queue(&dev->queues[0]). Here, we only check that the admin_q is non-NULL, rather than not an error or NULL, and begin quiescing a queue that never existed, leading to bad / NULL pointer dereference.
https://git.kernel.org/stable/c/f76729662650cd7bc8f8194e057af381370349a7
https://git.kernel.org/stable/c/da42761181627e9bdc37d18368b827948a583929
https://git.kernel.org/stable/c/af98940dd33c9f9e1beb4f71c0a39260100e2a65
https://git.kernel.org/stable/c/9e649471b396fa0139d53919354ce1eace9b9a24
https://git.kernel.org/stable/c/906c81dba8ee8057523859b5e1a2479e9fd34860
https://git.kernel.org/stable/c/8da2b7bdb47e94bbc4062a3978c708926bcb022c
https://git.kernel.org/stable/c/8321b17789f614414206af07e17ce4751c95dc76
https://git.kernel.org/stable/c/7a28556082d1fbcbc599baf1c24252dfc73efefc
https://git.kernel.org/stable/c/54a4c1e47d1b2585e74920399455bd9abbfb2bd7