Messages in this thread | | | Date | Wed, 13 Jan 2021 10:58:16 -0800 | From | Roman Gushchin <> | Subject | Re: [bpf] 755e5d5536: BUG:Bad_page_map_in_process |
| |
On Tue, Jan 12, 2021 at 09:27:51PM +0800, kernel test robot wrote: > > Greeting, > > FYI, we noticed the following commit (built with gcc-9): > > commit: 755e5d55367af5ff75a4db9b6cf439416878e2c7 ("bpf: Eliminate rlimit-based memory accounting for hashtab maps") > https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master > > > in testcase: trinity > version: trinity-i386-4d2343bd-1_20200320 > with following parameters: > > runtime: 300s > > test-description: Trinity is a linux system call fuzz tester. > test-url: http://codemonkey.org.uk/projects/trinity/ > > > on test machine: qemu-system-i386 -enable-kvm -cpu SandyBridge -smp 2 -m 8G > > caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace): > > > +------------------------------------------------------------+------------+------------+ > | | 844f157f6c | 755e5d5536 | > +------------------------------------------------------------+------------+------------+ > | BUG:Bad_page_map_in_process | 0 | 9 | > | BUG:Bad_page_state_in_process | 0 | 9 | > | BUG:Bad_rss-counter_state_mm:(ptrval)type:MM_FILEPAGES_val | 0 | 1 | > | BUG:Bad_rss-counter_state_mm:#type:MM_FILEPAGES_val | 0 | 8 | > | WARNING:at_mm/vmalloc.c:#vmap_pte_range | 0 | 3 | > | EIP:vmap_pte_range | 0 | 3 | > | BUG:unable_to_handle_page_fault_for_address | 0 | 1 | > | Oops:#[##] | 0 | 1 | > | EIP:free_percpu | 0 | 1 | > | EIP:__rb_reserve_next | 0 | 1 | > | WARNING:at_mm/percpu-vm.c:#__pcpu_balance_workfn | 0 | 1 | > | EIP:__pcpu_balance_workfn | 0 | 1 | > | WARNING:at_mm/vmalloc.c:#unmap_kernel_range_noflush | 0 | 1 | > | EIP:unmap_kernel_range_noflush | 0 | 1 | > +------------------------------------------------------------+------------+------------+ > > > If you fix the issue, kindly add following tag > Reported-by: kernel test robot <oliver.sang@intel.com>
I've tried to reproduce the problem on two different machines running the test case many times, but haven't succeed yet. I wonder though, if it's the same problem as described and fixed by commit e1868b9e36d0 ("bpf: Avoid overflows involving hash elem_size").
Thanks!
| |