From: Rasmus Villemoes <linux@rasmusvillemoes.dk>
To: linux-kernel@vger.kernel.org
Cc: x86@kernel.org, "H . Peter Anvin" <hpa@zytor.com>,
Ingo Molnar <mingo@kernel.org>,
"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>,
Rasmus Villemoes <linux@rasmusvillemoes.dk>
Subject: [POC 12/12] fs/dcache.c: use rai_bucket_shift for dentry hashtable
Date: Thu, 18 Oct 2018 00:33:32 +0200 [thread overview]
Message-ID: <20181017223332.11964-12-linux@rasmusvillemoes.dk> (raw)
In-Reply-To: <20181017223332.11964-1-linux@rasmusvillemoes.dk>
Before this, the disassembly of __d_lookup_rcu begins
0x2d10 <__d_lookup_rcu>: push %r15
0x2d12 <__d_lookup_rcu+2>: push %r14
0x2d14 <__d_lookup_rcu+4>: push %r13
0x2d16 <__d_lookup_rcu+6>: push %r12
0x2d18 <__d_lookup_rcu+8>: push %rbp
0x2d19 <__d_lookup_rcu+9>: push %rbx
0x2d1a <__d_lookup_rcu+10>: sub $0x18,%rsp
0x2d1e <__d_lookup_rcu+14>: mov (%rsi),%r12
0x2d21 <__d_lookup_rcu+17>: mov 0x0(%rip),%ecx # 0x2d27 <__d_lookup_rcu+23>
0x2d27 <__d_lookup_rcu+23>: mov 0x8(%rsi),%r13
0x2d2b <__d_lookup_rcu+27>: mov %r12d,%eax
0x2d2e <__d_lookup_rcu+30>: shr %cl,%eax
0x2d30 <__d_lookup_rcu+32>: mov 0x0(%rip),%rcx # 0x2d37 <__d_lookup_rcu+39>
0x2d37 <__d_lookup_rcu+39>: lea (%rcx,%rax,8),%rax
0x2d3b <__d_lookup_rcu+43>: mov (%rax),%rbx
0x2d3e <__d_lookup_rcu+46>: and $0xfffffffffffffffe,%rbx
0x2d42 <__d_lookup_rcu+50>: je 0x2df3 <__d_lookup_rcu+227>
0x2d48 <__d_lookup_rcu+56>: mov %r12,%rax
0x2d4b <__d_lookup_rcu+59>: mov %r12d,%r10d
0x2d4e <__d_lookup_rcu+62>: mov %rdx,%r15
After this, and after the patching of the run-time values of
dentry_hashtable and d_hash_shift, gdb on /proc/kcore says that we now
have
0xffffffff8902e8d0: push %r15
0xffffffff8902e8d2: push %r14
0xffffffff8902e8d4: push %r13
0xffffffff8902e8d6: push %r12
0xffffffff8902e8d8: push %rbp
0xffffffff8902e8d9: push %rbx
0xffffffff8902e8da: sub $0x18,%rsp
0xffffffff8902e8de: mov (%rsi),%r12
0xffffffff8902e8e1: mov 0x8(%rsi),%r13
0xffffffff8902e8e5: mov %r12d,%eax
0xffffffff8902e8e8: movabs $0xffff959b80007000,%rcx
0xffffffff8902e8f2: shr $0x12,%eax
0xffffffff8902e8f5: lea (%rcx,%rax,8),%rcx
0xffffffff8902e8f9: mov (%rcx),%rbx
0xffffffff8902e8fc: and $0xfffffffffffffffe,%rbx
0xffffffff8902e900: je 0xffffffff8902e9b1
0xffffffff8902e906: mov %r12,%rax
0xffffffff8902e909: mov %r12d,%r10d
0xffffffff8902e90c: mov %rdx,%r15
The shr $0x12, %eax is consistent with
[ 0.300676] Dentry cache hash table entries: 16384 (order: 5, 131072 bytes)
and nothing seems to explode.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
---
fs/dcache.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/dcache.c b/fs/dcache.c
index 1d54dfb38c9d..226298c3a599 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -100,7 +100,7 @@ static struct hlist_bl_head *dentry_hashtable __read_mostly;
static inline struct hlist_bl_head *d_hash(unsigned int hash)
{
- return dentry_hashtable + (hash >> d_hash_shift);
+ return rai_bucket_shift(dentry_hashtable, d_hash_shift, hash);
}
#define IN_LOOKUP_SHIFT 10
--
2.19.1.6.gbde171bbf5
prev parent reply other threads:[~2018-10-17 22:34 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-17 22:33 [POC 01/12] Accessing __ro_after_init variables as immediates Rasmus Villemoes
2018-10-17 22:33 ` [POC 02/12] init/main.c: call update_rai_access() Rasmus Villemoes
2018-10-17 22:33 ` [POC 03/12] arch/Kconfig: add ARCH_HAS_RAI symbol Rasmus Villemoes
2018-10-17 22:33 ` [POC 04/12] vmlinux.lds.h: handle various rai sections Rasmus Villemoes
2018-10-17 22:33 ` [POC 05/12] x86-64: initial ro-after-init patching support Rasmus Villemoes
2018-10-17 22:33 ` [POC 06/12] ugly ugly hack Rasmus Villemoes
2018-10-17 22:33 ` [POC 07/12] x86-64: rai: implement _rai_load Rasmus Villemoes
2018-10-17 22:33 ` [POC 08/12] fs/dcache.c: access dentry_cache via rai_load Rasmus Villemoes
2018-10-17 22:33 ` [POC 09/12] fs/inode.c: access inode_cachep " Rasmus Villemoes
2018-10-17 22:33 ` [POC 10/12] hack: /proc/rai: add rai_bucket_shift use Rasmus Villemoes
2018-10-17 22:33 ` [POC 11/12] x86-64: implement _rai_bucket_shift Rasmus Villemoes
2018-10-17 22:33 ` Rasmus Villemoes [this message]
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=20181017223332.11964-12-linux@rasmusvillemoes.dk \
--to=linux@rasmusvillemoes.dk \
--cc=hpa@zytor.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--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: 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®