From: Andy Lutomirski <luto@amacapital.net> To: Tony Luck <tony.luck@gmail.com> Cc: Dan Williams <dan.j.williams@intel.com>, Borislav Petkov <bp@alien8.de>, X86 ML <x86@kernel.org>, "linux-mm@kvack.org" <linux-mm@kvack.org>, Robert <elliott@hpe.com>, Andrew Morton <akpm@linux-foundation.org>, Ingo Molnar <mingo@kernel.org>, "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, linux-nvdimm <linux-nvdimm@ml01.01.org> Subject: Re: [PATCH v8 1/3] x86: Expand exception table to allow new handling options Date: Sat, 9 Jan 2016 14:32:43 -0800 [thread overview] Message-ID: <CALCETrUAO3gYiVpi5BO+o6=bika2D9JFZJ4xa9Ph8ArGMfftgA@mail.gmail.com> (raw) In-Reply-To: <CA+8MBbJHXTv=-OP1+dwq5KCursi8jRnWR5Mg=MavD_sVSY05eA@mail.gmail.com> On Jan 9, 2016 11:51 AM, "Tony Luck" <tony.luck@gmail.com> wrote: > > > Oh, I see. Is it the case that the MC code can't cleanly handle the > > case where the error was nominally recoverable but the kernel doesn't > > know how to recover from it due to the lack of a handler that's okay > > with it, because the handler's refusal to handle the fault wouldn't be > > known until too late? > > The code is just too clunky right now. We have a table driven > severity calculator that we invoke on each machine check bank > that has some valid data to report. Part of that calculation is > "what context am I in?". Which happens earlier in the sequence > than "Is MCi_STATUS.MCACOD some known recoverable type". > If I invoke the fixup code I'll change regs->ip right away ... even > if I'm executing on some innocent bystander processor that wasn't > the source of the machine check (the bystanders on the same > socket can usually see something logged in one of the memory > controller banks). Makes sense, sort of. But even if there is an MC fixup registered, don't you still have to make sure to execute it on the actual victim CPU? After all, you don't want to fail an mcsafe copy just because a different CPU coincidentally machine checked while the mcsafe copy has the recoverable RIP value. > > There are definitely some cleanups that should be done > in this code (e.g. figuring our context just once, not once > per bank). But I'm pretty sure I'll always want to know > "am I executing an instruction with a #MC recoverable > handler?" in a way that doesn't actually invoke the recovery. What's wrong with: Step 1: determine that the HW context is, in principle, recoverable. Step 2: ask the handler to try to recover. Step 3: if the handler doesn't recover, panic I'm not saying that restructuring the code like this should be a prerequisite for merging this, but I'm wondering whether it would make sense at some point in the future. --Andy
WARNING: multiple messages have this Message-ID (diff)
From: Andy Lutomirski <luto@amacapital.net> To: Tony Luck <tony.luck@gmail.com> Cc: Dan Williams <dan.j.williams@intel.com>, Borislav Petkov <bp@alien8.de>, X86 ML <x86@kernel.org>, "linux-mm@kvack.org" <linux-mm@kvack.org>, Robert <elliott@hpe.com>, Andrew Morton <akpm@linux-foundation.org>, Ingo Molnar <mingo@kernel.org>, "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, linux-nvdimm <linux-nvdimm@ml01.01.org> Subject: Re: [PATCH v8 1/3] x86: Expand exception table to allow new handling options Date: Sat, 9 Jan 2016 14:32:43 -0800 [thread overview] Message-ID: <CALCETrUAO3gYiVpi5BO+o6=bika2D9JFZJ4xa9Ph8ArGMfftgA@mail.gmail.com> (raw) In-Reply-To: <CA+8MBbJHXTv=-OP1+dwq5KCursi8jRnWR5Mg=MavD_sVSY05eA@mail.gmail.com> On Jan 9, 2016 11:51 AM, "Tony Luck" <tony.luck@gmail.com> wrote: > > > Oh, I see. Is it the case that the MC code can't cleanly handle the > > case where the error was nominally recoverable but the kernel doesn't > > know how to recover from it due to the lack of a handler that's okay > > with it, because the handler's refusal to handle the fault wouldn't be > > known until too late? > > The code is just too clunky right now. We have a table driven > severity calculator that we invoke on each machine check bank > that has some valid data to report. Part of that calculation is > "what context am I in?". Which happens earlier in the sequence > than "Is MCi_STATUS.MCACOD some known recoverable type". > If I invoke the fixup code I'll change regs->ip right away ... even > if I'm executing on some innocent bystander processor that wasn't > the source of the machine check (the bystanders on the same > socket can usually see something logged in one of the memory > controller banks). Makes sense, sort of. But even if there is an MC fixup registered, don't you still have to make sure to execute it on the actual victim CPU? After all, you don't want to fail an mcsafe copy just because a different CPU coincidentally machine checked while the mcsafe copy has the recoverable RIP value. > > There are definitely some cleanups that should be done > in this code (e.g. figuring our context just once, not once > per bank). But I'm pretty sure I'll always want to know > "am I executing an instruction with a #MC recoverable > handler?" in a way that doesn't actually invoke the recovery. What's wrong with: Step 1: determine that the HW context is, in principle, recoverable. Step 2: ask the handler to try to recover. Step 3: if the handler doesn't recover, panic I'm not saying that restructuring the code like this should be a prerequisite for merging this, but I'm wondering whether it would make sense at some point in the future. --Andy -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2016-01-09 22:33 UTC|newest] Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top 2016-01-09 0:04 [PATCH v8 0/3] Machine check recovery when kernel accesses poison Tony Luck 2016-01-09 0:04 ` Tony Luck 2015-12-31 19:40 ` [PATCH v8 2/3] x86, mce: Check for faults tagged in EXTABLE_CLASS_FAULT exception table entries Tony Luck 2015-12-31 19:40 ` Tony Luck 2016-01-08 20:49 ` [PATCH v8 1/3] x86: Expand exception table to allow new handling options Tony Luck 2016-01-08 20:49 ` Tony Luck 2016-01-09 1:52 ` Andy Lutomirski 2016-01-09 1:52 ` Andy Lutomirski 2016-01-09 3:39 ` Brian Gerst 2016-01-09 3:39 ` Brian Gerst 2016-01-09 4:31 ` Brian Gerst 2016-01-09 4:31 ` Brian Gerst 2016-01-09 6:36 ` Andy Lutomirski 2016-01-09 6:36 ` Andy Lutomirski 2016-01-11 23:09 ` Brian Gerst 2016-01-11 23:09 ` Brian Gerst 2016-01-11 23:22 ` Andy Lutomirski 2016-01-11 23:22 ` Andy Lutomirski 2016-01-11 23:48 ` Luck, Tony 2016-01-11 23:48 ` Luck, Tony 2016-01-09 17:45 ` Tony Luck 2016-01-09 17:45 ` Tony Luck 2016-01-09 18:00 ` Andy Lutomirski 2016-01-09 18:00 ` Andy Lutomirski 2016-01-09 19:51 ` Tony Luck 2016-01-09 19:51 ` Tony Luck 2016-01-09 22:32 ` Andy Lutomirski [this message] 2016-01-09 22:32 ` Andy Lutomirski 2016-01-10 1:15 ` Tony Luck 2016-01-10 1:15 ` Tony Luck 2016-01-11 0:25 ` Luck, Tony 2016-01-11 0:25 ` Luck, Tony 2016-01-08 21:18 ` [PATCH v8 3/3] x86, mce: Add __mcsafe_copy() Tony Luck 2016-01-08 21:18 ` Tony Luck 2016-01-09 1:49 ` Andy Lutomirski 2016-01-09 1:49 ` Andy Lutomirski 2016-01-09 17:48 ` Tony Luck 2016-01-09 17:48 ` Tony Luck 2016-01-09 17:57 ` Andy Lutomirski 2016-01-09 17:57 ` Andy Lutomirski 2016-01-09 19:39 ` Tony Luck 2016-01-09 19:39 ` Tony Luck 2016-01-09 22:15 ` Dan Williams 2016-01-09 22:15 ` Dan Williams 2016-01-09 22:33 ` Andy Lutomirski 2016-01-09 22:33 ` Andy Lutomirski 2016-01-10 0:23 ` Dan Williams 2016-01-10 0:23 ` Dan Williams 2016-01-10 1:40 ` Tony Luck 2016-01-10 1:40 ` Tony Luck 2016-01-10 11:26 ` Borislav Petkov 2016-01-10 11:26 ` Borislav Petkov 2016-01-11 10:44 ` Ingo Molnar 2016-01-11 10:44 ` Ingo Molnar 2016-01-13 23:22 ` Tony Luck 2016-01-13 23:22 ` Tony Luck 2016-01-14 4:39 ` Borislav Petkov 2016-01-14 4:39 ` Borislav Petkov 2016-01-30 0:35 ` Tony Luck 2016-01-30 0:35 ` Tony Luck 2016-01-30 10:28 ` Borislav Petkov 2016-01-30 10:28 ` Borislav Petkov 2016-02-01 23:10 ` Tony Luck 2016-02-01 23:10 ` Tony Luck 2016-02-01 23:16 ` Dan Williams 2016-02-01 23:16 ` Dan Williams 2016-01-12 0:26 ` Luck, Tony 2016-01-12 0:26 ` Luck, Tony 2016-01-12 0:30 ` Andy Lutomirski 2016-01-12 0:30 ` Andy Lutomirski 2016-01-12 0:37 ` Andy Lutomirski 2016-01-12 0:37 ` Andy Lutomirski
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to='CALCETrUAO3gYiVpi5BO+o6=bika2D9JFZJ4xa9Ph8ArGMfftgA@mail.gmail.com' \ --to=luto@amacapital.net \ --cc=akpm@linux-foundation.org \ --cc=bp@alien8.de \ --cc=dan.j.williams@intel.com \ --cc=elliott@hpe.com \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=linux-nvdimm@ml01.01.org \ --cc=mingo@kernel.org \ --cc=tony.luck@gmail.com \ --cc=x86@kernel.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.