Re: [PATCH v2] uart: fix race between uart_put_char() and uart_shutdown()
On Fri, Jun 29, 2018 at 10:43:30AM -0600, Tycho Andersen wrote: > On Fri, Jun 29, 2018 at 04:24:46AM -0600, Tycho Andersen wrote: > > v2: switch to locking uport->lock on allocation/deallocation instead of > > locking the per-port mutex in uart_put_char. Note that since > > uport->lock is a spin lock, we have to switch the allocation to > > GFP_ATOMIC. > > Serge pointed out off-list that we may want to do the allocation > before the lock so that it's more likely to be successful. I'm happy > to send that change to this if it's what we want to do, I don't have a > strong preference. That sounds like a much better thing to do. thanks, greg k-h
Re: [PATCH v2] uart: fix race between uart_put_char() and uart_shutdown()
On Fri, Jun 29, 2018 at 10:43:30AM -0600, Tycho Andersen wrote: > On Fri, Jun 29, 2018 at 04:24:46AM -0600, Tycho Andersen wrote: > > v2: switch to locking uport->lock on allocation/deallocation instead of > > locking the per-port mutex in uart_put_char. Note that since > > uport->lock is a spin lock, we have to switch the allocation to > > GFP_ATOMIC. > > Serge pointed out off-list that we may want to do the allocation > before the lock so that it's more likely to be successful. I'm happy > to send that change to this if it's what we want to do, I don't have a > strong preference. That sounds like a much better thing to do. thanks, greg k-h
Re: [PATCH v2] uart: fix race between uart_put_char() and uart_shutdown()
On Fri, Jun 29, 2018 at 04:24:46AM -0600, Tycho Andersen wrote: > v2: switch to locking uport->lock on allocation/deallocation instead of > locking the per-port mutex in uart_put_char. Note that since > uport->lock is a spin lock, we have to switch the allocation to > GFP_ATOMIC. Serge pointed out off-list that we may want to do the allocation before the lock so that it's more likely to be successful. I'm happy to send that change to this if it's what we want to do, I don't have a strong preference. Tycho
Re: [PATCH v2] uart: fix race between uart_put_char() and uart_shutdown()
On Fri, Jun 29, 2018 at 04:24:46AM -0600, Tycho Andersen wrote: > v2: switch to locking uport->lock on allocation/deallocation instead of > locking the per-port mutex in uart_put_char. Note that since > uport->lock is a spin lock, we have to switch the allocation to > GFP_ATOMIC. Serge pointed out off-list that we may want to do the allocation before the lock so that it's more likely to be successful. I'm happy to send that change to this if it's what we want to do, I don't have a strong preference. Tycho
[PATCH v2] uart: fix race between uart_put_char() and uart_shutdown()
We have reports of the following crash: PID: 7 TASK: 88085c6d61c0 CPU: 1 COMMAND: "kworker/u25:0" #0 [88085c6db710] machine_kexec at 81046239 #1 [88085c6db760] crash_kexec at 810fc248 #2 [88085c6db830] oops_end at 81008ae7 #3 [88085c6db860] no_context at 81050b8f #4 [88085c6db8b0] __bad_area_nosemaphore at 81050d75 #5 [88085c6db900] bad_area_nosemaphore at 81050e83 #6 [88085c6db910] __do_page_fault at 8105132e #7 [88085c6db9b0] do_page_fault at 8105152c #8 [88085c6db9c0] page_fault at 81a3f122 [exception RIP: uart_put_char+149] RIP: 814b67b5 RSP: 88085c6dba78 RFLAGS: 00010006 RAX: 0292 RBX: 827c5120 RCX: 0081 RDX: RSI: 005f RDI: 827c5120 RBP: 88085c6dba98 R8: 012c R9: 822ea320 R10: 88085fe4db04 R11: 0001 R12: 881059f9c000 R13: 0001 R14: 005f R15: 0fba ORIG_RAX: CS: 0010 SS: 0018 #9 [88085c6dbaa0] tty_put_char at 81497544 #10 [88085c6dbac0] do_output_char at 8149c91c #11 [88085c6dbae0] __process_echoes at 8149cb8b #12 [88085c6dbb30] commit_echoes at 8149cdc2 #13 [88085c6dbb60] n_tty_receive_buf_fast at 8149e49b #14 [88085c6dbbc0] __receive_buf at 8149ef5a #15 [88085c6dbc20] n_tty_receive_buf_common at 8149f016 #16 [88085c6dbca0] n_tty_receive_buf2 at 8149f194 #17 [88085c6dbcb0] flush_to_ldisc at 814a238a #18 [88085c6dbd50] process_one_work at 81090be2 #19 [88085c6dbe20] worker_thread at 81091b4d #20 [88085c6dbeb0] kthread at 81096384 #21 [88085c6dbf50] ret_from_fork at 81a3d69f after slogging through some dissasembly: 814b6720 : 814b6720: 55 push %rbp 814b6721: 48 89 e5mov%rsp,%rbp 814b6724: 48 83 ec 20 sub$0x20,%rsp 814b6728: 48 89 1c 24 mov%rbx,(%rsp) 814b672c: 4c 89 64 24 08 mov%r12,0x8(%rsp) 814b6731: 4c 89 6c 24 10 mov%r13,0x10(%rsp) 814b6736: 4c 89 74 24 18 mov%r14,0x18(%rsp) 814b673b: e8 b0 8e 58 00 callq 81a3f5f0 814b6740: 4c 8b a7 88 02 00 00mov0x288(%rdi),%r12 814b6747: 45 31 edxor%r13d,%r13d 814b674a: 41 89 f6mov%esi,%r14d 814b674d: 49 83 bc 24 70 01 00cmpq $0x0,0x170(%r12) 814b6754: 00 00 814b6756: 49 8b 9c 24 80 01 00mov0x180(%r12),%rbx 814b675d: 00 814b675e: 74 2f je 814b678f 814b6760: 48 89 dfmov%rbx,%rdi 814b6763: e8 a8 67 58 00 callq 81a3cf10 <_raw_spin_lock_irqsave> 814b6768: 41 8b 8c 24 78 01 00mov0x178(%r12),%ecx 814b676f: 00 814b6770: 89 ca mov%ecx,%edx 814b6772: f7 d2 not%edx 814b6774: 41 03 94 24 7c 01 00add0x17c(%r12),%edx 814b677b: 00 814b677c: 81 e2 ff 0f 00 00 and$0xfff,%edx 814b6782: 75 23 jne814b67a7 814b6784: 48 89 c6mov%rax,%rsi 814b6787: 48 89 dfmov%rbx,%rdi 814b678a: e8 e1 64 58 00 callq 81a3cc70 <_raw_spin_unlock_irqrestore> 814b678f: 44 89 e8mov%r13d,%eax 814b6792: 48 8b 1c 24 mov(%rsp),%rbx 814b6796: 4c 8b 64 24 08 mov0x8(%rsp),%r12 814b679b: 4c 8b 6c 24 10 mov0x10(%rsp),%r13 814b67a0: 4c 8b 74 24 18 mov0x18(%rsp),%r14 814b67a5: c9 leaveq 814b67a6: c3 retq 814b67a7: 49 8b 94 24 70 01 00mov0x170(%r12),%rdx 814b67ae: 00 814b67af: 48 63 c9movslq %ecx,%rcx 814b67b2: 41 b5 01mov$0x1,%r13b 814b67b5: 44 88 34 0a mov%r14b,(%rdx,%rcx,1) 814b67b9: 41 8b 94 24 78 01 00mov0x178(%r12),%edx 814b67c0: 00 814b67c1: 83 c2 01add$0x1,%edx 814b67c4: 81 e2 ff 0f 00 00 and$0xfff,%edx 814b67ca: 41 89 94 24 78 01 00mov%edx,0x178(%r12) 814b67d1:
[PATCH v2] uart: fix race between uart_put_char() and uart_shutdown()
We have reports of the following crash: PID: 7 TASK: 88085c6d61c0 CPU: 1 COMMAND: "kworker/u25:0" #0 [88085c6db710] machine_kexec at 81046239 #1 [88085c6db760] crash_kexec at 810fc248 #2 [88085c6db830] oops_end at 81008ae7 #3 [88085c6db860] no_context at 81050b8f #4 [88085c6db8b0] __bad_area_nosemaphore at 81050d75 #5 [88085c6db900] bad_area_nosemaphore at 81050e83 #6 [88085c6db910] __do_page_fault at 8105132e #7 [88085c6db9b0] do_page_fault at 8105152c #8 [88085c6db9c0] page_fault at 81a3f122 [exception RIP: uart_put_char+149] RIP: 814b67b5 RSP: 88085c6dba78 RFLAGS: 00010006 RAX: 0292 RBX: 827c5120 RCX: 0081 RDX: RSI: 005f RDI: 827c5120 RBP: 88085c6dba98 R8: 012c R9: 822ea320 R10: 88085fe4db04 R11: 0001 R12: 881059f9c000 R13: 0001 R14: 005f R15: 0fba ORIG_RAX: CS: 0010 SS: 0018 #9 [88085c6dbaa0] tty_put_char at 81497544 #10 [88085c6dbac0] do_output_char at 8149c91c #11 [88085c6dbae0] __process_echoes at 8149cb8b #12 [88085c6dbb30] commit_echoes at 8149cdc2 #13 [88085c6dbb60] n_tty_receive_buf_fast at 8149e49b #14 [88085c6dbbc0] __receive_buf at 8149ef5a #15 [88085c6dbc20] n_tty_receive_buf_common at 8149f016 #16 [88085c6dbca0] n_tty_receive_buf2 at 8149f194 #17 [88085c6dbcb0] flush_to_ldisc at 814a238a #18 [88085c6dbd50] process_one_work at 81090be2 #19 [88085c6dbe20] worker_thread at 81091b4d #20 [88085c6dbeb0] kthread at 81096384 #21 [88085c6dbf50] ret_from_fork at 81a3d69f after slogging through some dissasembly: 814b6720 : 814b6720: 55 push %rbp 814b6721: 48 89 e5mov%rsp,%rbp 814b6724: 48 83 ec 20 sub$0x20,%rsp 814b6728: 48 89 1c 24 mov%rbx,(%rsp) 814b672c: 4c 89 64 24 08 mov%r12,0x8(%rsp) 814b6731: 4c 89 6c 24 10 mov%r13,0x10(%rsp) 814b6736: 4c 89 74 24 18 mov%r14,0x18(%rsp) 814b673b: e8 b0 8e 58 00 callq 81a3f5f0 814b6740: 4c 8b a7 88 02 00 00mov0x288(%rdi),%r12 814b6747: 45 31 edxor%r13d,%r13d 814b674a: 41 89 f6mov%esi,%r14d 814b674d: 49 83 bc 24 70 01 00cmpq $0x0,0x170(%r12) 814b6754: 00 00 814b6756: 49 8b 9c 24 80 01 00mov0x180(%r12),%rbx 814b675d: 00 814b675e: 74 2f je 814b678f 814b6760: 48 89 dfmov%rbx,%rdi 814b6763: e8 a8 67 58 00 callq 81a3cf10 <_raw_spin_lock_irqsave> 814b6768: 41 8b 8c 24 78 01 00mov0x178(%r12),%ecx 814b676f: 00 814b6770: 89 ca mov%ecx,%edx 814b6772: f7 d2 not%edx 814b6774: 41 03 94 24 7c 01 00add0x17c(%r12),%edx 814b677b: 00 814b677c: 81 e2 ff 0f 00 00 and$0xfff,%edx 814b6782: 75 23 jne814b67a7 814b6784: 48 89 c6mov%rax,%rsi 814b6787: 48 89 dfmov%rbx,%rdi 814b678a: e8 e1 64 58 00 callq 81a3cc70 <_raw_spin_unlock_irqrestore> 814b678f: 44 89 e8mov%r13d,%eax 814b6792: 48 8b 1c 24 mov(%rsp),%rbx 814b6796: 4c 8b 64 24 08 mov0x8(%rsp),%r12 814b679b: 4c 8b 6c 24 10 mov0x10(%rsp),%r13 814b67a0: 4c 8b 74 24 18 mov0x18(%rsp),%r14 814b67a5: c9 leaveq 814b67a6: c3 retq 814b67a7: 49 8b 94 24 70 01 00mov0x170(%r12),%rdx 814b67ae: 00 814b67af: 48 63 c9movslq %ecx,%rcx 814b67b2: 41 b5 01mov$0x1,%r13b 814b67b5: 44 88 34 0a mov%r14b,(%rdx,%rcx,1) 814b67b9: 41 8b 94 24 78 01 00mov0x178(%r12),%edx 814b67c0: 00 814b67c1: 83 c2 01add$0x1,%edx 814b67c4: 81 e2 ff 0f 00 00 and$0xfff,%edx 814b67ca: 41 89 94 24 78 01 00mov%edx,0x178(%r12) 814b67d1:
