Re: "BUG: using smp_processor_id() in preemptible" with KPTI on 4.14.11 - Greg Kroah-Hartman

From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: Thomas Zeitlhofer <thomas.zeitlhofer+lkml@ze-it.at>,
	Thomas Gleixner <tglx@linutronix.de>
Cc: Hugh Dickins <hughd@google.com>, linux-kernel@vger.kernel.org
Subject: Re: "BUG: using smp_processor_id() in preemptible" with KPTI on 4.14.11
Date: Thu, 4 Jan 2018 13:55:28 +0100	[thread overview]
Message-ID: <20180104125528.GA15548@kroah.com> (raw)
In-Reply-To: <20180104124320.eawuo6q7wnwzpf7s@toau>

On Thu, Jan 04, 2018 at 01:43:20PM +0100, Thomas Zeitlhofer wrote:
> On Thu, Jan 04, 2018 at 11:51:11AM +0100, Greg Kroah-Hartman wrote:
> > On Thu, Jan 04, 2018 at 11:20:29AM +0100, Thomas Zeitlhofer wrote:
> > > On Thu, Jan 04, 2018 at 02:59:06AM +0100, Thomas Zeitlhofer wrote:
> > > > Hello,
> > > > 
> > > > on an Ivybridge CPU, I get with 4.14.11:
> > > > 
> > > >    BUG: using smp_processor_id() in preemptible [00000000] code: ovsdb-server/4510
> > > >    caller is native_flush_tlb_single+0x57/0xc0
> > > >    CPU: 3 PID: 4510 Comm: ovsdb-server Not tainted 4.14.11-kvm-00434-gcd0b8eb84f5c #3
> > > >    Hardware name: MSI MS-7798/B75MA-P45 (MS-7798), BIOS V1.9 09/30/2013
> > > >    Call Trace:
> > > >     dump_stack+0x5c/0x86
> > > >     check_preemption_disabled+0xdd/0xe0
> > > >     native_flush_tlb_single+0x57/0xc0
> > > >     ? __set_pte_vaddr+0x2d/0x40
> > > >     __set_pte_vaddr+0x2d/0x40
> > > >     set_pte_vaddr+0x2f/0x40
> > > >     cea_set_pte+0x30/0x40
> > > >     ds_update_cea.constprop.4+0x4d/0x70
> > > >     reserve_ds_buffers+0x159/0x410
> > > >     ? wp_page_copy+0x36d/0x6a0
> > > >     x86_reserve_hardware+0x150/0x160
> > > >     x86_pmu_event_init+0x3e/0x1f0
> > > >     perf_try_init_event+0x69/0x80
> > > >     perf_event_alloc+0x652/0x740
> > > >     SyS_perf_event_open+0x3f6/0xd60
> > > >     do_syscall_64+0x5c/0x190
> > > >     entry_SYSCALL64_slow_path+0x25/0x25
> > > >    RIP: 0033:0x74a1d94580b9
> > > >    RSP: 002b:00007fff0c01d5d8 EFLAGS: 00000206 ORIG_RAX: 000000000000012a
> > > >    RAX: ffffffffffffffda RBX: 00007fff0c01d7b0 RCX: 000074a1d94580b9
> > > >    RDX: 00000000ffffffff RSI: 0000000000000000 RDI: 00007fff0c01d5e0
> > > >    RBP: 0000000000000000 R08: 0000000000000000 R09: 0000007000000000
> > > >    R10: 00000000ffffffff R11: 0000000000000206 R12: 0000000000000008
> > > >    R13: 0000000000000000 R14: 00007fff0c01d790 R15: 00005df43a799600
> > > > 
> > > > This does not show up when booting with pti=off.
> > > > 
> > > > Maybe it is related to the issue that is fixed for the upcoming 4.4.110
> > > > release by https://lkml.org/lkml/2018/1/3/692
> > 
> > I don't understand this link.  
> 
> I found that link when trying to search for the error message. That
> patch touches __native_flush_tlb_single() and mentions hardware
> differences in Ivybridge and below:
> 
> 	"We have many machines (Westmere, Sandybridge, Ivybridge)
> 	supporting PCID but not INVPCID..."
> 
> As I see the error message only on Ivybridge and not on Haswell, I came
> up with the vague guess that this could be related.
> 
> > The 4.4 and 4.9 backports are much different than the 4.14 tree.
> 
> Yes, I have seen that.
> 
> > > JFYI, the very same kernel does not show this issue on a Haswell CPU.
> > 
> > I have now queued up a bunch of patches that are in Linus's tree, can
> > you test these out as well:
> > 	https://git.kernel.org/pub/scm/linux/kernel/git/stable/stable-queue.git/tree/queue-4.14
> 
> Does not seem to make any difference - with those patches applied I
> still get:
> 
>    BUG: using smp_processor_id() in preemptible [00000000] code: ovsdb-server/4383
>    caller is native_flush_tlb_single+0x57/0xc0
>    CPU: 3 PID: 4383 Comm: ovsdb-server Not tainted 4.14.11-kvm-00435-g3138001170c9 #3
>    Hardware name: MSI MS-7798/B75MA-P45 (MS-7798), BIOS V1.9 09/30/2013
>    Call Trace:
>     dump_stack+0x5c/0x86
>     check_preemption_disabled+0xdd/0xe0
>     native_flush_tlb_single+0x57/0xc0
>     ? __set_pte_vaddr+0x2d/0x40
>     __set_pte_vaddr+0x2d/0x40
>     set_pte_vaddr+0x2f/0x40
>     cea_set_pte+0x30/0x40
>     ds_update_cea.constprop.4+0x4d/0x70
>     reserve_ds_buffers+0x159/0x410
>     ? wp_page_copy+0x36d/0x6a0
>     x86_reserve_hardware+0x150/0x160
>     x86_pmu_event_init+0x3e/0x1f0
>     perf_try_init_event+0x69/0x80
>     perf_event_alloc+0x652/0x740
>     SyS_perf_event_open+0x3f6/0xd60
>     do_syscall_64+0x5c/0x190
>     entry_SYSCALL64_slow_path+0x25/0x25
>    RIP: 0033:0x755c0b8580b9
>    RSP: 002b:00007fffc87cf9e8 EFLAGS: 00000206 ORIG_RAX: 000000000000012a
>    RAX: ffffffffffffffda RBX: 00007fffc87cfbc0 RCX: 0000755c0b8580b9
>    RDX: 00000000ffffffff RSI: 0000000000000000 RDI: 00007fffc87cf9f0
>    RBP: 0000000000000000 R08: 0000000000000000 R09: 0000007000000000
>    R10: 00000000ffffffff R11: 0000000000000206 R12: 0000000000000008
>    R13: 0000000000000000 R14: 00007fffc87cfba0 R15: 000062ea2cbff600
> 

Odd, does 4.15-rc6 also trigger the same error?  Thomas is working on an
issue with KALSR (see lkml with:
	Subject: Re: "bad pmd" errors + oops with KPTI on 4.14.11 after loading X.509 certs
)

thanks,

greg k-h