linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Helge Deller <deller@gmx.de>
Subject: [PATCH 4.8 07/33] parisc: Fix TLB related boot crash on SMP machines
Date: Tue, 13 Dec 2016 09:16:08 -0800	[thread overview]
Message-ID: <20161213171534.467335471@linuxfoundation.org> (raw)
In-Reply-To: <20161213171534.171564506@linuxfoundation.org>

4.8-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Helge Deller <deller@gmx.de>

commit 24d0492b7d5d321a9c5846c8c974eba9823ffaa0 upstream.

At bootup we run measurements to calculate the best threshold for when we
should be using full TLB flushes instead of just flushing a specific amount of
TLB entries.  This performance test is run over the kernel text segment.

But running this TLB performance test on the kernel text segment turned out to
crash some SMP machines when the kernel text pages were mapped as huge pages.

To avoid those crashes this patch simply skips this test on some SMP machines
and calculates an optimal threshold based on the maximum number of available
TLB entries and number of online CPUs.

On a technical side, this seems to happen:
The TLB measurement code uses flush_tlb_kernel_range() to flush specific TLB
entries with a page size of 4k (pdtlb 0(sr1,addr)). On UP systems this purge
instruction seems to work without problems even if the pages were mapped as
huge pages.  But on SMP systems the TLB purge instruction is broadcasted to
other CPUs. Those CPUs then crash the machine because the page size is not as
expected.  C8000 machines with PA8800/PA8900 CPUs were not affected by this
problem, because the required cache coherency prohibits to use huge pages at
all.  Sadly I didn't found any documentation about this behaviour, so this
finding is purely based on testing with phyiscal SMP machines (A500-44 and
J5000, both were 2-way boxes).

Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 arch/parisc/kernel/cache.c |   11 +++++++++++
 1 file changed, 11 insertions(+)

--- a/arch/parisc/kernel/cache.c
+++ b/arch/parisc/kernel/cache.c
@@ -393,6 +393,15 @@ void __init parisc_setup_cache_timing(vo
 
 	/* calculate TLB flush threshold */
 
+	/* On SMP machines, skip the TLB measure of kernel text which
+	 * has been mapped as huge pages. */
+	if (num_online_cpus() > 1 && !parisc_requires_coherency()) {
+		threshold = max(cache_info.it_size, cache_info.dt_size);
+		threshold *= PAGE_SIZE;
+		threshold /= num_online_cpus();
+		goto set_tlb_threshold;
+	}
+
 	alltime = mfctl(16);
 	flush_tlb_all();
 	alltime = mfctl(16) - alltime;
@@ -411,6 +420,8 @@ void __init parisc_setup_cache_timing(vo
 		alltime, size, rangetime);
 
 	threshold = PAGE_ALIGN(num_online_cpus() * size * alltime / rangetime);
+
+set_tlb_threshold:
 	if (threshold)
 		parisc_tlb_flush_threshold = threshold;
 	printk(KERN_INFO "TLB flush threshold set to %lu KiB\n",

  parent reply	other threads:[~2016-12-13 17:21 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20161213171629epcas1p4dc2737fb27df6f1d6148edf348dde1c7@epcas1p4.samsung.com>
2016-12-13 17:16 ` [PATCH 4.8 00/33] 4.8.15-stable review Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 01/33] powerpc/eeh: Fix deadlock when PE frozen state cant be cleared Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 02/33] powerpc/mm: Fix lazy icache flush on pre-POWER5 Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 03/33] powerpc/boot: Fix build failure in 32-bit boot wrapper Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 04/33] fuse: fix clearing suid, sgid for chown() Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 05/33] parisc: Purge TLB before setting PTE Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 06/33] parisc: Remove unnecessary TLB purges from flush_dcache_page_asm and flush_icache_page_asm Greg Kroah-Hartman
2016-12-13 17:16   ` Greg Kroah-Hartman [this message]
2016-12-13 17:16   ` [PATCH 4.8 08/33] zram: restrict add/remove attributes to root only Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 09/33] locking/rtmutex: Prevent dequeue vs. unlock race Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 10/33] locking/rtmutex: Use READ_ONCE() in rt_mutex_owner() Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 11/33] device-dax: fix private mapping restriction, permit read-only Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 12/33] scsi: lpfc: fix oops/BUG in lpfc_sli_ringtxcmpl_put() Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 13/33] sched/autogroup: Fix 64-bit kernel nice level adjustment Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 14/33] vhost-vsock: fix orphan connection reset Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 15/33] perf/x86: Fix full width counter, counter overflow Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 16/33] acpi, nfit: fix extended status translations for ACPI DSMs Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 17/33] acpi, nfit, libnvdimm: fix / harden ars_status output length handling Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 18/33] acpi, nfit: validate ars_status output buffer size Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 19/33] acpi, nfit: fix bus vs dimm confusion in xlat_status Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 20/33] crypto: marvell - Dont copy hash operation twice into the SRAM Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 22/33] crypto: mcryptd - Check mcryptd algorithm compatibility Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 23/33] crypto: marvell - Dont corrupt state of an STD req for re-stepped ahash Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 24/33] can: raw: raw_setsockopt: limit number of can_filter that can be set Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 26/33] Revert "ACPI: Execute _PTS before system reboot" Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 27/33] ARM: dts: orion5x: fix number of sata port for linkstation ls-gl Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 28/33] ARM: dts: imx7d: fix LCDIF clock assignment Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 29/33] ceph: dont set req->r_locked_dir in ceph_d_revalidate Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 30/33] m68k: Fix ndelay() macro Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 31/33] batman-adv: Check for alloc errors when preparing TT local data Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 32/33] hotplug: Make register and unregister notifier API symmetric Greg Kroah-Hartman
2016-12-13 17:16   ` [PATCH 4.8 33/33] crypto: rsa - Add Makefile dependencies to fix parallel builds Greg Kroah-Hartman
2016-12-13 19:32   ` [PATCH 4.8 00/33] 4.8.15-stable review Shuah Khan
2016-12-13 20:07     ` Greg Kroah-Hartman
2016-12-14  4:13   ` Guenter Roeck
2016-12-14 12:49     ` Greg Kroah-Hartman

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=20161213171534.467335471@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=deller@gmx.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.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: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).