linux-kernel - Re: [PATCH] x86/hyperv: Fix hv_get/set

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 15 Feb 2023 22:05:29 +0530
From:   Jinank Jain <jinankjain@...ux.microsoft.com>
To:     Wei Liu <wei.liu@...nel.org>,
        Nuno Das Neves <nunodasneves@...ux.microsoft.com>
Cc:     linux-hyperv@...r.kernel.org, linux-kernel@...r.kernel.org,
        x86@...nel.org, mikelley@...rosoft.com, kys@...rosoft.com,
        Tianyu.Lan@...rosoft.com, haiyangz@...rosoft.com,
        decui@...rosoft.com, tglx@...utronix.de, mingo@...hat.com,
        bp@...en8.de, dave.hansen@...ux.intel.com, hpa@...or.com
Subject: Re: [PATCH] x86/hyperv: Fix hv_get/set_register for nested bringup

The patch looks good to me, apart from the comments from Wei regarding 
styling.

On 2/13/2023 7:58 PM, Wei Liu wrote:
> A few comments on style.
>
> On Thu, Feb 09, 2023 at 02:02:52PM -0800, Nuno Das Neves wrote:
>> hv_get_nested_reg only translates SINT0, resulting in the wrong sint
>> being registered by nested vmbus.
> Please put a blank line between paragraphs.
>
>> Fix the issue with new utility function hv_is_sint_reg.
>> While at it, improve clarity of hv_set_non_nested_register and hv_is_synic_reg.
>>
>> Signed-off-by: Nuno Das Neves <nunodasneves@...ux.microsoft.com>
>> ---
>>   arch/x86/include/asm/mshyperv.h | 11 +++++++----
>>   arch/x86/kernel/cpu/mshyperv.c  |  8 ++++----
>>   2 files changed, 11 insertions(+), 8 deletions(-)
>>
>> diff --git a/arch/x86/include/asm/mshyperv.h b/arch/x86/include/asm/mshyperv.h
>> index 9ae1a344536b..684c547c1cca 100644
>> --- a/arch/x86/include/asm/mshyperv.h
>> +++ b/arch/x86/include/asm/mshyperv.h
>> @@ -225,10 +225,13 @@ extern bool hv_isolation_type_snp(void);
>>   
>>   static inline bool hv_is_synic_reg(unsigned int reg)
>>   {
>> -	if ((reg >= HV_REGISTER_SCONTROL) &&
>> -	    (reg <= HV_REGISTER_SINT15))
>> -		return true;
>> -	return false;
>> +	return (reg >= HV_REGISTER_SCONTROL) &&
>> +	       (reg <= HV_REGISTER_SINT15);
>> +}
> Please put a new line here.
>
> I can fix these issues too if you don't end up sending a new version due
> to other issues.
>
> Jinank, please take a look. The code looks sensible to me, but I would
> like you to have a look too.
>
> Thanks,
> Wei.
>
>> +static inline bool hv_is_sint_reg(unsigned int reg)
>> +{
>> +	return (reg >= HV_REGISTER_SINT0) &&
>> +	       (reg <= HV_REGISTER_SINT15);
>>   }
>>   
>>   u64 hv_get_register(unsigned int reg);
>> diff --git a/arch/x86/kernel/cpu/mshyperv.c b/arch/x86/kernel/cpu/mshyperv.c
>> index 0ceb6d1f9c3c..6bd344e1200f 100644
>> --- a/arch/x86/kernel/cpu/mshyperv.c
>> +++ b/arch/x86/kernel/cpu/mshyperv.c
>> @@ -44,6 +44,9 @@ struct ms_hyperv_info ms_hyperv;
>>   #if IS_ENABLED(CONFIG_HYPERV)
>>   static inline unsigned int hv_get_nested_reg(unsigned int reg)
>>   {
>> +	if (hv_is_sint_reg(reg))
>> +		return reg - HV_REGISTER_SINT0 + HV_REGISTER_NESTED_SINT0;
>> +
>>   	switch (reg) {
>>   	case HV_REGISTER_SIMP:
>>   		return HV_REGISTER_NESTED_SIMP;
>> @@ -53,8 +56,6 @@ static inline unsigned int hv_get_nested_reg(unsigned int reg)
>>   		return HV_REGISTER_NESTED_SVERSION;
>>   	case HV_REGISTER_SCONTROL:
>>   		return HV_REGISTER_NESTED_SCONTROL;
>> -	case HV_REGISTER_SINT0:
>> -		return HV_REGISTER_NESTED_SINT0;
>>   	case HV_REGISTER_EOM:
>>   		return HV_REGISTER_NESTED_EOM;
>>   	default:
>> @@ -80,8 +81,7 @@ void hv_set_non_nested_register(unsigned int reg, u64 value)
>>   		hv_ghcb_msr_write(reg, value);
>>   
>>   		/* Write proxy bit via wrmsl instruction */
>> -		if (reg >= HV_REGISTER_SINT0 &&
>> -		    reg <= HV_REGISTER_SINT15)
>> +		if (hv_is_sint_reg(reg))
>>   			wrmsrl(reg, value | 1 << 20);
>>   	} else {
>>   		wrmsrl(reg, value);
>> -- 
>> 2.25.1

Reviewed-by: Jinank Jain <jinankjain@...ux.microsoft.com>