From: Alex Shi <alex.shi@intel.com>
To: tglx@linutronix.de, mingo@redhat.com, hpa@zytor.com,
arnd@arndb.de, rostedt@goodmis.org, fweisbec@gmail.com
Cc: jeremy@goop.org, seto.hidetoshi@jp.fujitsu.com,
borislav.petkov@amd.com, alex.shi@intel.com, tony.luck@intel.com,
luto@mit.edu, riel@redhat.com, avi@redhat.com,
len.brown@intel.com, dhowells@redhat.com, yinghai@kernel.org,
ak@linux.intel.com, jbeulich@suse.com, akpm@linux-foundation.org,
eric.dumazet@gmail.com, akinobu.mita@gmail.com, cpw@sgi.com,
steiner@sgi.com, penberg@kernel.org, a.p.zijlstra@chello.nl,
hughd@google.com, kamezawa.hiroyu@jp.fujitsu.com,
viro@zeniv.linux.org.uk, linux-kernel@vger.kernel.org,
yongjie.ren@intel.com
Subject: [PATCH v7 6/8] x86/tlb: add tlb_flushall_shift knob into debugfs
Date: Wed, 23 May 2012 22:15:53 +0800 [thread overview]
Message-ID: <1337782555-8088-7-git-send-email-alex.shi@intel.com> (raw)
In-Reply-To: <1337782555-8088-1-git-send-email-alex.shi@intel.com>
kernel will replace cr3 rewrite with invlpg when
tlb_flush_entries <= active_tlb_entries / 2^tlb_flushall_factor
if tlb_flushall_factor is -1, kernel won't do this replacement.
User can modify its value according to specific CPU/applications.
Thanks for Borislav providing the help message of
CONFIG_DEBUG_TLBFLUSH.
Signed-off-by: Alex Shi <alex.shi@intel.com>
---
arch/x86/Kconfig.debug | 19 +++++++++++++++++
arch/x86/mm/tlb.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 70 insertions(+), 0 deletions(-)
diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index e46c214..b322f12 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -129,6 +129,25 @@ config DOUBLEFAULT
option saves about 4k and might cause you much additional grey
hair.
+config DEBUG_TLBFLUSH
+ bool "Set upper limit of TLB entries to flush one-by-one"
+ depends on DEBUG_KERNEL && (X86_64 || X86_INVLPG)
+ ---help---
+
+ X86-only for now.
+
+ This option allows the user to tune the amount of TLB entries the
+ kernel flushes one-by-one instead of doing a full TLB flush. In
+ certain situations, the former is cheaper. This is controlled by the
+ tlb_flushall_shift knob under /sys/kernel/debug/x86. If you set it
+ to -1, the code flushes the whole TLB unconditionally. Otherwise,
+ for positive values of it, the kernel will use single TLB entry
+ invalidating instructions according to the following formula:
+
+ flush_entries <= active_tlb_entries / 2^tlb_flushall_shift
+
+ If in doubt, say "N".
+
config IOMMU_DEBUG
bool "Enable IOMMU debugging"
depends on GART_IOMMU && DEBUG_KERNEL
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index d0ab80e..0f17b04 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -12,6 +12,7 @@
#include <asm/cache.h>
#include <asm/apic.h>
#include <asm/uv/uv.h>
+#include <linux/debugfs.h>
DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate)
= { &init_mm, 0, };
@@ -421,3 +422,53 @@ void flush_tlb_all(void)
{
on_each_cpu(do_flush_tlb_all, NULL, 1);
}
+
+#ifdef CONFIG_DEBUG_TLBFLUSH
+static ssize_t tlbflush_read_file(struct file *file, char __user *user_buf,
+ size_t count, loff_t *ppos)
+{
+ char buf[32];
+ unsigned int len;
+
+ len = sprintf(buf, "%hd\n", tlb_flushall_shift);
+ return simple_read_from_buffer(user_buf, count, ppos, buf, len);
+}
+
+static ssize_t tlbflush_write_file(struct file *file,
+ const char __user *user_buf, size_t count, loff_t *ppos)
+{
+ char buf[32];
+ ssize_t len;
+ s8 shift;
+
+ len = min(count, sizeof(buf) - 1);
+ if (copy_from_user(buf, user_buf, len))
+ return -EFAULT;
+
+ buf[len] = '\0';
+ if (kstrtos8(buf, 0, &shift))
+ return -EINVAL;
+
+ if (shift > 64)
+ return -EINVAL;
+
+ tlb_flushall_shift = shift;
+ return count;
+}
+
+static const struct file_operations fops_tlbflush = {
+ .read = tlbflush_read_file,
+ .write = tlbflush_write_file,
+ .llseek = default_llseek,
+};
+
+static int __cpuinit create_tlb_flushall_shift(void)
+{
+ if (cpu_has_invlpg) {
+ debugfs_create_file("tlb_flushall_shift", S_IRUSR | S_IWUSR,
+ arch_debugfs_dir, NULL, &fops_tlbflush);
+ }
+ return 0;
+}
+late_initcall(create_tlb_flushall_shift);
+#endif
--
1.7.5.4
next prev parent reply other threads:[~2012-05-23 14:17 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-05-23 14:15 [PATCH v7 0/8] x86 tlb optimisations Alex Shi
2012-05-23 14:15 ` [PATCH v7 1/8] x86/tlb_info: get last level TLB entry number of CPU Alex Shi
2012-05-23 14:15 ` [PATCH v7 2/8] x86/flush_tlb: try flush_tlb_single one by one in flush_tlb_range Alex Shi
2012-05-23 14:51 ` Jan Beulich
2012-05-24 6:41 ` Alex Shi
2012-05-24 8:12 ` Jan Beulich
2012-05-24 8:55 ` Alex Shi
2012-05-24 9:44 ` Jan Beulich
2012-05-24 14:36 ` Alex Shi
2012-05-25 2:43 ` Alex Shi
2012-05-23 14:15 ` [PATCH v7 3/8] x86/tlb: fall back to flush all when meet a THP large page Alex Shi
2012-05-23 14:15 ` [PATCH v7 4/8] x86/tlb: add tlb_flushall_shift for specific CPU Alex Shi
2012-05-23 14:15 ` [PATCH v7 5/8] x86/tlb: enable tlb flush range support for generic mmu and x86 Alex Shi
2012-05-23 14:15 ` Alex Shi [this message]
2012-05-23 14:15 ` [PATCH v7 7/8] x86/tlb: replace INVALIDATE_TLB_VECTOR by CALL_FUNCTION_VECTOR Alex Shi
2012-05-23 14:15 ` [PATCH v7 8/8] x86/tlb: just do tlb flush on one of siblings of SMT Alex Shi
2012-05-23 15:05 ` Jan Beulich
2012-05-23 17:09 ` Peter Zijlstra
2012-05-23 17:15 ` Peter Zijlstra
2012-05-24 1:46 ` Andrew Lutomirski
2012-05-24 5:12 ` Alex Shi
2012-05-24 6:04 ` Borislav Petkov
2012-05-24 7:40 ` Peter Zijlstra
2012-05-24 13:19 ` Andrew Lutomirski
2012-05-24 13:23 ` Peter Zijlstra
2012-05-24 13:39 ` Arjan van de Ven
2012-05-24 13:54 ` Alex Shi
2012-05-24 14:18 ` Arjan van de Ven
2012-05-24 14:32 ` Alex Shi
2012-05-24 15:03 ` H. Peter Anvin
2012-05-25 0:24 ` Alex Shi
2012-05-24 16:08 ` Arjan van de Ven
2012-05-25 0:28 ` Alex Shi
2012-05-25 0:46 ` Arjan van de Ven
2012-05-24 8:32 ` Alex Shi
2012-05-24 8:42 ` Peter Zijlstra
2012-05-24 8:48 ` Alex Shi
2012-05-24 11:35 ` Rusty Russell
2012-05-24 14:03 ` Alex Shi
2012-05-24 9:27 ` Alex Shi
2012-05-24 9:42 ` Peter Zijlstra
2012-05-24 9:46 ` Jan Beulich
2012-05-24 14:06 ` Alex Shi
2012-05-24 8:43 ` Peter Zijlstra
2012-05-24 8:48 ` Jan Beulich
2012-05-24 9:02 ` Alex Shi
2012-05-24 9:45 ` Jan Beulich
2012-05-24 15:04 ` H. Peter Anvin
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=1337782555-8088-7-git-send-email-alex.shi@intel.com \
--to=alex.shi@intel.com \
--cc=a.p.zijlstra@chello.nl \
--cc=ak@linux.intel.com \
--cc=akinobu.mita@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=avi@redhat.com \
--cc=borislav.petkov@amd.com \
--cc=cpw@sgi.com \
--cc=dhowells@redhat.com \
--cc=eric.dumazet@gmail.com \
--cc=fweisbec@gmail.com \
--cc=hpa@zytor.com \
--cc=hughd@google.com \
--cc=jbeulich@suse.com \
--cc=jeremy@goop.org \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=len.brown@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@mit.edu \
--cc=mingo@redhat.com \
--cc=penberg@kernel.org \
--cc=riel@redhat.com \
--cc=rostedt@goodmis.org \
--cc=seto.hidetoshi@jp.fujitsu.com \
--cc=steiner@sgi.com \
--cc=tglx@linutronix.de \
--cc=tony.luck@intel.com \
--cc=viro@zeniv.linux.org.uk \
--cc=yinghai@kernel.org \
--cc=yongjie.ren@intel.com \
/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
all inboxes | Powered by JetHome®