Received-SPF: pass (google.com: domain of linux-kernel+bounces-186263-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Message-ID: <4a090901-9705-40aa-ac3d-d67c52660f22@amd.com>
Date: Wed, 22 May 2024 18:42:55 +0530
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH v2 1/1] x86/elf: Add a new .note section containing
 Xfeatures information to x86 core files
Content-Language: en-US
To: Thomas Gleixner <tglx@linutronix.de>,
 "Balasubrmanian, Vignesh" <Vignesh.Balasubrmanian@amd.com>,
 "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
 "linux-toolchains@vger.kernel.org" <linux-toolchains@vger.kernel.org>
Cc: "mpe@ellerman.id.au" <mpe@ellerman.id.au>,
 "npiggin@gmail.com" <npiggin@gmail.com>,
 "christophe.leroy@csgroup.eu" <christophe.leroy@csgroup.eu>,
 "aneesh.kumar@kernel.org" <aneesh.kumar@kernel.org>,
 "naveen.n.rao@linux.ibm.com" <naveen.n.rao@linux.ibm.com>,
 "ebiederm@xmission.com" <ebiederm@xmission.com>,
 "keescook@chromium.org" <keescook@chromium.org>,
 "x86@kernel.org" <x86@kernel.org>,
 "linuxppc-dev@lists.ozlabs.org" <linuxppc-dev@lists.ozlabs.org>,
 "linux-mm@kvack.org" <linux-mm@kvack.org>,
 "Petkov, Borislav" <Borislav.Petkov@amd.com>,
 "George, Jini Susan" <JiniSusan.George@amd.com>, "matz@suse.de"
 <matz@suse.de>, "binutils@sourceware.org" <binutils@sourceware.org>,
 "jhb@FreeBSD.org" <jhb@FreeBSD.org>,
 "felix.willgerodt@intel.com" <felix.willgerodt@intel.com>
References: <87wmo4o3r4.ffs@tglx>
From: "Balasubrmanian, Vignesh" <vigbalas@amd.com>
In-Reply-To: <87wmo4o3r4.ffs@tglx>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0:
	=?utf-8?B?RllPTjN0V0gvc1diYVRpL1hTdjE5akRKSSt6ZW5MdUtCNTdKdzNYeGQzNCtl?=
 =?utf-8?B?N1VEOWIzOVl1SDB2ckY0R3llS2xlbGNEa2R2L0xNb3REb1JnaUFyNVZiVWlQ?=
 =?utf-8?B?d3AvZ1ZvMG0xVVlvYmNMNkhlQlRMOXU2M0tIMVIwaUFuRThQZHY0Z1d6NnpO?=
 =?utf-8?B?WThGTVRtaXhGSjZXTWptT3psV1o4OXNHaVQ4M1Z0dHY5ZnZzWGZTb21YSW5z?=
 =?utf-8?B?S0VMWU1ZVm1UWjE2MjZWRGtWZG91dXZtaVlkVXl2VFZaL016ZzhGbHYwNnRU?=
 =?utf-8?B?bGM3VTFWU29HYWhaQXU5Mjd0Y1NCYmN3d284OHFJeXJDcW1kTVpYVForaDZH?=
 =?utf-8?B?Ri9pWno5aSt0cm1mcWlpQ2k0UDMvN3UyTGlpTTJzanh3WnpWcUJ4Mkl4dmRS?=
 =?utf-8?B?T2F2RWVkeGREeHhyRFhaYjhNQ1lpRi9YcStMMGdHOTRQUnRKcG1xNldmTXpS?=
 =?utf-8?B?dVNxSGwxVlF5Szh4OUJWcnhGREM2cUZtaWhKRUVnR2pJVjdHS3dmYmc2RUlu?=
 =?utf-8?B?TWc3WkdwUml1MlkyMkR4R3N2NHBVeGJxelZHbmVkclhuOHNoK2pxRnFjZ0Rz?=
 =?utf-8?B?dTc1ek5VZkhVR0c5QnZMYVQrZm9aMitnQkhzWHZsM2JVMmNDRys2UjBwZHBN?=
 =?utf-8?B?WEUrKzZ2YmdNVGxTcklGcVhKd2h6eDdSREg4TVRPSnlyRDc1MllYck8zYnpK?=
 =?utf-8?B?c1hkbXI3UHF6Tjd1KzUrYm55bGNrcjRYY0sxRUVRNnFnaUdtcXNRdG1xM1RW?=
 =?utf-8?B?RTRHZjMxWjd2aSt5cHk5N3Z2dmt2M2Z5a1hEc1dKRW1ZbjliYUNRNUpzQitS?=
 =?utf-8?B?VUFFR3ZxajBzY3JJd2FIQ1loZUVPZ3NTVGRCR1Rvb2NSK1VIMGJZSU5waFYz?=
 =?utf-8?B?MXFQOFhZczBDZ21HOXVmaVVKTVo5WGtvb2NnaG9BVEY1ektwT0RpNklhNTlJ?=
 =?utf-8?B?RkJCclh1QWsxeEJyU21ZeDhqMFk1TTZiT0Y3RFFzYy81MmxERnRtcmZoSEJx?=
 =?utf-8?B?eDdtekFZaGpKcGtqelk1OUFvWTllMW9IWk9qalMzaDMyeWdWNTJCbHhvTzA3?=
 =?utf-8?B?OXZyYVRNaTBMY0xkSHBYQlp2ek9oZ0htNG5PUDlUVDVaUnAyYW9TR3pHdHlJ?=
 =?utf-8?B?QlZMUmp4aE5YSXVPYndLenRSNUJSY1d2ZXYvWEYrOGhyWWEwdDBGUEhuNXRP?=
 =?utf-8?B?TXJydkhMRUV5bmdmVFQ5bG0xT1ZpRWtMa1NObDk4aXQzWUp2RDNIdnlxNGx5?=
 =?utf-8?B?eEI2RS9wWXh4MmtlRkpQRW8yeEt6WEV4MDk2MExVWW44TzFpL05aL2Q3VHVr?=
 =?utf-8?B?QmdKRThrZXdLN2dCMnlLMkNBSXQzMFVnZ0JsUGRBK3BsWm9mQVRwcEpvZllK?=
 =?utf-8?B?Nk1NbmtVTzN0Zm4vbnBHSGIzTGVSUDI0WE9Sc0pUaHBvQnlSalAyVjNqcXJK?=
 =?utf-8?B?L1NpYTdZaFJZbFlnV0srQ3F0SGlHMEwwd3VLMEFnQ1JMb1dCNU5zbmdEanE4?=
 =?utf-8?B?NVpJMW5ZdlZEbmdWMUZmSHlYV1JXSklCVFM1cXRLMy9BeDRNMUFZQmxXbUl5?=
 =?utf-8?B?bDVQNkRzb3NwUmo1K3AwSzVFNzc4SzV0SGFVY0psU2JBOERvYjBSKzVOaDlo?=
 =?utf-8?B?WktibElxczkvUEltOFFUUjZiVGRxZzVld0Y0K1lqWlRsYTk3cnJmRzRkUGlh?=
 =?utf-8?B?YjFtd0NEc3lHZ1d0K01hOFZjbXRJeHRYSUt6OThXR0MyTkJDeTdFNWdvQkdO?=
 =?utf-8?B?VzBjNHowck1tOXhXZHk3Z1lUUjdTK0RERVBsRmZKOTNwV2UraDFuazIrZW8r?=
 =?utf-8?B?OTlNMmRydjFNTnh1dTMzSGk4ang3K3lDZVZYWUszTStMeVcxME1LditqUUJR?=
 =?utf-8?B?R0dNc0FLYmh0Z09BZ0NtdGxKNWJnSWtuQVhyTXErK2VlU3JjL09ERTN4c3Za?=
 =?utf-8?B?WTNQNkpGaXNFVWpkc2lDYUpRa3VvWStVZGVEREhrTjRURDhTWnRqbUFqekI0?=
 =?utf-8?B?bm9uQVVIVDdsTlNFTEVPZFEzdnUwa2V2SmZjOFhMYjFzUXF2VngrN0I5VEJE?=
 =?utf-8?B?dk5nd3lOeDFRb21IZ1RwWW9Mbmc1V0xhY0g0R0t5QUZuMFkwa090ajNBUGZI?=
 =?utf-8?Q?6ezgMczFJs47ofQTsMT6FJHsJ?=
X-OriginatorOrg: amd.com
X-MS-Exchange-CrossTenant-Network-Message-Id: d8a7ecb2-75c7-46f6-e9b0-08dc7a60eb37
X-MS-Exchange-CrossTenant-AuthSource: SJ1PR12MB6124.namprd12.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 22 May 2024 13:13:06.3932
 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 3dd8961f-e488-4e60-8e11-a82d994e183d
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: rHLmuNIms05OKwKvU4kzy3LDb+VfhM96ajCvghw/mUPQCP0feguVQsI+x6/woe4DJaAfOl0Hu9afkNri6ClgfQ==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: PH0PR12MB7011


On 5/8/2024 6:32 PM, Thomas Gleixner wrote:
> Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.
>
>
> On Tue, May 07 2024 at 15:23, Vignesh Balasubramanian wrote:
>> +struct xfeat_component {
>> +     u32 xfeat_type;
>> +     u32 xfeat_sz;
>> +     u32 xfeat_off;
>> +     u32 xfeat_flags;
>> +} __packed;
> Why repeating xfeat_ for all member names?
>
>      u32       type;
>      u32       size;
>      u32       offset;
>      u32       flags;
>
> is sufficient and obvious, no?
>
>> +enum custom_feature {
>> +     FEATURE_XSAVE_FP = 0,
>> +     FEATURE_XSAVE_SSE = 1,
>> +     FEATURE_XSAVE_YMM = 2,
>> +     FEATURE_XSAVE_BNDREGS = 3,
>> +     FEATURE_XSAVE_BNDCSR = 4,
>> +     FEATURE_XSAVE_OPMASK = 5,
>> +     FEATURE_XSAVE_ZMM_Hi256 = 6,
>> +     FEATURE_XSAVE_Hi16_ZMM = 7,
>> +     FEATURE_XSAVE_PT = 8,
>> +     FEATURE_XSAVE_PKRU = 9,
>> +     FEATURE_XSAVE_PASID = 10,
>> +     FEATURE_XSAVE_CET_USER = 11,
>> +     FEATURE_XSAVE_CET_SHADOW_STACK = 12,
>> +     FEATURE_XSAVE_HDC = 13,
>> +     FEATURE_XSAVE_UINTR = 14,
>> +     FEATURE_XSAVE_LBR = 15,
>> +     FEATURE_XSAVE_HWP = 16,
>> +     FEATURE_XSAVE_XTILE_CFG = 17,
>> +     FEATURE_XSAVE_XTILE_DATA = 18,
>> +     FEATURE_MAX,
>> +     FEATURE_XSAVE_EXTENDED_START = FEATURE_XSAVE_YMM,
>> +     FEATURE_XSAVE_EXTENDED_END = FEATURE_XSAVE_XTILE_DATA,
>> +};
> Why can't this use the existing 'enum xfeature' which is providing
> exactly the same information already?
First version of patch was similar to what you mentioned here and other 
review comments to use existing kernel definitions.
https://lore.kernel.org/linux-mm/20240314112359.50713-1-vigbalas@amd.com/T/

As per the review comment 
https://lore.kernel.org/linux-mm/20240314162954.GAZfMmAnYQoRjRbRzc@fat_crate.local/ 
, modified the patch to be a independent of kernel internal definitions.
Though this enum and below function  "get_sub_leaf" are not useful now,  
it will be required when we extend for a new/different features.

Please let  us know your suggestions.

I will fix all other review comments in my next version.

>> +#ifdef CONFIG_COREDUMP
>> +static int get_sub_leaf(int custom_xfeat)
>> +{
>> +     switch (custom_xfeat) {
>> +     case FEATURE_XSAVE_YMM:                 return XFEATURE_YMM;
>> +     case FEATURE_XSAVE_BNDREGS:             return XFEATURE_BNDREGS;
>> +     case FEATURE_XSAVE_BNDCSR:              return XFEATURE_BNDCSR;
>> +     case FEATURE_XSAVE_OPMASK:              return XFEATURE_OPMASK;
>> +     case FEATURE_XSAVE_ZMM_Hi256:           return XFEATURE_ZMM_Hi256;
>> +     case FEATURE_XSAVE_Hi16_ZMM:            return XFEATURE_Hi16_ZMM;
>> +     case FEATURE_XSAVE_PT:                  return XFEATURE_PT_UNIMPLEMENTED_SO_FAR;
>> +     case FEATURE_XSAVE_PKRU:                return XFEATURE_PKRU;
>> +     case FEATURE_XSAVE_PASID:               return XFEATURE_PASID;
>> +     case FEATURE_XSAVE_CET_USER:            return XFEATURE_CET_USER;
>> +     case FEATURE_XSAVE_CET_SHADOW_STACK:    return XFEATURE_CET_KERNEL_UNUSED;
>> +     case FEATURE_XSAVE_HDC:                 return XFEATURE_RSRVD_COMP_13;
>> +     case FEATURE_XSAVE_UINTR:               return XFEATURE_RSRVD_COMP_14;
>> +     case FEATURE_XSAVE_LBR:                 return XFEATURE_LBR;
>> +     case FEATURE_XSAVE_HWP:                 return XFEATURE_RSRVD_COMP_16;
>> +     case FEATURE_XSAVE_XTILE_CFG:           return XFEATURE_XTILE_CFG;
>> +     case FEATURE_XSAVE_XTILE_DATA:          return XFEATURE_XTILE_DATA;
>> +     default:
>> +             pr_warn_ratelimited("Not a valid XSAVE Feature.");
>> +             return 0;
>> +     }
>> +}
> This function then maps the identical enums one to one. The only actual
> "functionality" is the default case and that's completely pointless.
thanks,
vigneshbalu.