[anolis-intel-cloud:devel-5.10 33/33] arch/x86/kernel/cpu/bugs.c:1379:20: warning: this statement may fall through

[kernel test robot <lkp@intel.com>]

tree:   https://gitee.com/anolis/intel-cloud-kernel.git devel-5.10
head:   ff3fd7c43fb0c95eb699767b7b4052b0250c87ce
commit: ff3fd7c43fb0c95eb699767b7b4052b0250c87ce [33/33] anolis: cpu vulnerabilities: support hardware-based mitigation for Retbleed and SRSO via Hygon IBRS to avoid performance degradation caused by software-based mitigation.
config: x86_64-anolis-intel-cloud-devel-5.10 (https://download.01.org/0day-ci/archive/20260115/202601151910.FPutxmgQ-lkp@intel.com/config)
compiler: gcc-14 (Debian 14.2.0-19) 14.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260115/202601151910.FPutxmgQ-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202601151910.FPutxmgQ-lkp@intel.com/

All warnings (new ones prefixed by >>):

   arch/x86/kernel/cpu/bugs.c: In function 'spectre_v2_determine_rsb_fill_type_at_vmexit':
>> arch/x86/kernel/cpu/bugs.c:1379:20: warning: this statement may fall through [-Wimplicit-fallthrough=]
    1379 |                 if (boot_cpu_data.x86_vendor == X86_VENDOR_HYGON)
         |                    ^
   arch/x86/kernel/cpu/bugs.c:1381:9: note: here
    1381 |         case SPECTRE_V2_RETPOLINE:
         |         ^~~~


vim +1379 arch/x86/kernel/cpu/bugs.c

  1342	
  1343	static void __init spectre_v2_determine_rsb_fill_type_at_vmexit(enum spectre_v2_mitigation mode)
  1344	{
  1345		/*
  1346		 * Similar to context switches, there are two types of RSB attacks
  1347		 * after VM exit:
  1348		 *
  1349		 * 1) RSB underflow
  1350		 *
  1351		 * 2) Poisoned RSB entry
  1352		 *
  1353		 * When retpoline is enabled, both are mitigated by filling/clearing
  1354		 * the RSB.
  1355		 *
  1356		 * When IBRS is enabled, while #1 would be mitigated by the IBRS branch
  1357		 * prediction isolation protections, RSB still needs to be cleared
  1358		 * because of #2.  Note that SMEP provides no protection here, unlike
  1359		 * user-space-poisoned RSB entries.
  1360		 *
  1361		 * eIBRS should protect against RSB poisoning, but if the EIBRS_PBRSB
  1362		 * bug is present then a LITE version of RSB protection is required,
  1363		 * just a single call needs to retire before a RET is executed.
  1364		 */
  1365		switch (mode) {
  1366		case SPECTRE_V2_NONE:
  1367			return;
  1368	
  1369		case SPECTRE_V2_EIBRS_LFENCE:
  1370		case SPECTRE_V2_EIBRS:
  1371			if (boot_cpu_has_bug(X86_BUG_EIBRS_PBRSB)) {
  1372				setup_force_cpu_cap(X86_FEATURE_RSB_VMEXIT_LITE);
  1373				pr_info("Spectre v2 / PBRSB-eIBRS: Retire a single CALL on VMEXIT\n");
  1374			}
  1375			return;
  1376	
  1377		case SPECTRE_V2_EIBRS_RETPOLINE:
  1378			/* Hygon Enhanced IBRS flushes RAS upon privilege level changes from low to high. */
> 1379			if (boot_cpu_data.x86_vendor == X86_VENDOR_HYGON)
  1380				return;
  1381		case SPECTRE_V2_RETPOLINE:
  1382		case SPECTRE_V2_LFENCE:
  1383		case SPECTRE_V2_IBRS:
  1384			setup_force_cpu_cap(X86_FEATURE_RSB_VMEXIT);
  1385			pr_info("Spectre v2 / SpectreRSB : Filling RSB on VMEXIT\n");
  1386			return;
  1387		}
  1388	
  1389		pr_warn_once("Unknown Spectre v2 mode, disabling RSB mitigation at VM exit");
  1390		dump_stack();
  1391	}
  1392	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki