From: Rusty Russell <rusty@rustcorp.com.au>
To: Andrew Morton <akpm@osdl.org>
Cc: matthieu castet <castet.matthieu@free.fr>,
lkml - Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [patch] bug in cpuid & msr on nosmp machine
Date: Thu, 20 May 2004 18:34:36 +1000 [thread overview]
Message-ID: <1085042076.7541.27.camel@bach> (raw)
In-Reply-To: <20040520003240.75fd355d.akpm@osdl.org>
On Thu, 2004-05-20 at 17:32, Andrew Morton wrote:
> I think what you want here is
>
> if (!cpu_possible(cpu) || !cpu_online(cpu))
> return -ENXIO;
It works, but it's not really correct. cpu_possible() is correct, but
cpu_online() might no longer be true by the time do_cpu_read() calls
do_cpu_id().
One way would be to do lock_cpu_hotplug() in cpuid_open() and introduce
a cpuid_close() which would do unlock_cpu_hotplug(). Another would be
to drop the check here, and fail the actual read. A final way is to do
no checks, in which case it becomes a noop if the cpu is offline.
> --- 25/arch/i386/kernel/cpuid.c~cpuid-msr-range-checking-fix 2004-05-20 00:30:21.812166544 -0700
> +++ 25-akpm/arch/i386/kernel/cpuid.c 2004-05-20 00:31:16.607836336 -0700
> @@ -133,10 +133,12 @@ static ssize_t cpuid_read(struct file *f
> static int cpuid_open(struct inode *inode, struct file *file)
> {
> int cpu = iminor(file->f_dentry->d_inode);
> - struct cpuinfo_x86 *c = &(cpu_data)[cpu];
> + struct cpuinfo_x86 *c;
>
> - if (!cpu_online(cpu))
> + if (!cpu_possible(cpu) || !cpu_online(cpu))
> return -ENXIO; /* No such CPU */
> +
> + c = &(cpu_data)[cpu];
> if (c->cpuid_level < 0)
> return -EIO; /* CPUID not supported */
>
> diff -puN arch/i386/kernel/msr.c~cpuid-msr-range-checking-fix arch/i386/kernel/msr.c
> --- 25/arch/i386/kernel/msr.c~cpuid-msr-range-checking-fix 2004-05-20 00:30:21.836162896 -0700
> +++ 25-akpm/arch/i386/kernel/msr.c 2004-05-20 00:31:56.952702984 -0700
> @@ -239,10 +239,12 @@ static ssize_t msr_write(struct file * f
> static int msr_open(struct inode *inode, struct file *file)
> {
> int cpu = iminor(file->f_dentry->d_inode);
> - struct cpuinfo_x86 *c = &(cpu_data)[cpu];
> -
> - if (!cpu_online(cpu))
> - return -ENXIO; /* No such CPU */
> + struct cpuinfo_x86 *c;
> +
> + if (!cpu_possible(cpu) || !cpu_online(cpu))
> + return -ENXIO; /* No such CPU */
> +
> + c = &(cpu_data)[cpu];
> if ( !cpu_has(c, X86_FEATURE_MSR) )
> return -EIO; /* MSR not supported */
>
>
> _
--
Anyone who quotes me in their signature is an idiot -- Rusty Russell
next prev parent reply other threads:[~2004-05-20 8:35 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-05-19 16:35 [patch] bug in cpuid & msr on nosmp machine matthieu castet
2004-05-19 17:46 ` matthieu castet
2004-05-20 2:44 ` Jeff Sipek
2004-05-20 9:51 ` Paul Jackson
2004-05-20 7:32 ` Andrew Morton
2004-05-20 8:34 ` Rusty Russell [this message]
2004-05-20 8:51 ` Andrew Morton
2004-05-20 9:30 ` Rusty Russell
2004-05-20 9:57 ` Rusty Russell
[not found] <1XCh4-1jO-67@gated-at.bofh.it>
[not found] ` <1XRg3-4LW-31@gated-at.bofh.it>
[not found] ` <1XRg3-4LW-29@gated-at.bofh.it>
[not found] ` <1XSc3-5y3-23@gated-at.bofh.it>
2004-05-22 8:13 ` Andi Kleen
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=1085042076.7541.27.camel@bach \
--to=rusty@rustcorp.com.au \
--cc=akpm@osdl.org \
--cc=castet.matthieu@free.fr \
--cc=linux-kernel@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