Re: [PATCH] x86/mtrr: Drop workaround for old 32bit CPUs

From: Andrew Cooper <andrew.cooper3@citrix.com>
To: "Roger Pau Monné" <roger.pau@citrix.com>
Cc: Xen-devel <xen-devel@lists.xenproject.org>, Wei Liu <wl@xen.org>,
	Jan Beulich <JBeulich@suse.com>
Subject: Re: [PATCH] x86/mtrr: Drop workaround for old 32bit CPUs
Date: Wed, 8 Jul 2020 11:52:57 +0100	[thread overview]
Message-ID: <12874bc4-39e8-5ed4-3893-79154a206293@citrix.com> (raw)
In-Reply-To: <20200708104826.GB7191@Air-de-Roger>

On 08/07/2020 11:48, Roger Pau Monné wrote:
> On Wed, Jul 08, 2020 at 11:14:43AM +0100, Andrew Cooper wrote:
>> This logic is dead as Xen is 64bit-only these days.
>>
>> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
> Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>

Thanks,

>> ---
>> CC: Jan Beulich <JBeulich@suse.com>
>> CC: Wei Liu <wl@xen.org>
>> CC: Roger Pau Monné <roger.pau@citrix.com>
>> ---
>>  xen/arch/x86/cpu/mtrr/generic.c | 17 -----------------
>>  1 file changed, 17 deletions(-)
>>
>> diff --git a/xen/arch/x86/cpu/mtrr/generic.c b/xen/arch/x86/cpu/mtrr/generic.c
>> index 89634f918f..06fa0c0420 100644
>> --- a/xen/arch/x86/cpu/mtrr/generic.c
>> +++ b/xen/arch/x86/cpu/mtrr/generic.c
>> @@ -570,23 +570,6 @@ int generic_validate_add_page(unsigned long base, unsigned long size, unsigned i
>>  {
>>  	unsigned long lbase, last;
>>  
>> -	/*  For Intel PPro stepping <= 7, must be 4 MiB aligned 
>> -	    and not touch 0x70000000->0x7003FFFF */
>> -	if (is_cpu(INTEL) && boot_cpu_data.x86 == 6 &&
>> -	    boot_cpu_data.x86_model == 1 &&
>> -	    boot_cpu_data.x86_mask <= 7) {
>> -		if (base & ((1 << (22 - PAGE_SHIFT)) - 1)) {
>> -			printk(KERN_WARNING "mtrr: base(%#lx000) is not 4 MiB aligned\n", base);
>> -			return -EINVAL;
>> -		}
>> -		if (!(base + size < 0x70000 || base > 0x7003F) &&
>> -		    (type == MTRR_TYPE_WRCOMB
>> -		     || type == MTRR_TYPE_WRBACK)) {
>> -			printk(KERN_WARNING "mtrr: writable mtrr between 0x70000000 and 0x7003FFFF may hang the CPU.\n");
>> -			return -EINVAL;
>> -		}
>> -	}
>> -
>>  	/*  Check upper bits of base and last are equal and lower bits are 0
>>  	    for base and 1 for last  */
>>  	last = base + size - 1;
> FWIW, you could also initialize last at definition time.

I've got some very different cleanup in mind for that code, seeing as it
can be simplified to a single test expression.

~Andrew