Messages in this thread | | | Date | Tue, 9 Aug 2022 15:04:31 +0800 | From | kernel test robot <> | Subject | [intel-amx:init_fpstate 2/3] arch/x86/kernel/fpu/xstate.c:362:24: warning: format specifies type 'unsigned int' but the argument has type 'unsigned long' |
| |
tree: https://github.com/intel/amx-linux.git init_fpstate head: b1323896658f2b069193a782265a47e3c99611e4 commit: 86878d20f388dc243b534c20a5de1bedfb534df1 [2/3] x86/fpu: Avoid the init XSTATE buffer overflow config: x86_64-randconfig-r023-20220808 (https://download.01.org/0day-ci/archive/20220809/202208091405.uYHLXk68-lkp@intel.com/config) compiler: clang version 16.0.0 (https://github.com/llvm/llvm-project 5f1c7e2cc5a3c07cbc2412e851a7283c1841f520) reproduce (this is a W=1 build): wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross # https://github.com/intel/amx-linux/commit/86878d20f388dc243b534c20a5de1bedfb534df1 git remote add intel-amx https://github.com/intel/amx-linux.git git fetch --no-tags intel-amx init_fpstate git checkout 86878d20f388dc243b534c20a5de1bedfb534df1 # save the config file mkdir build_dir && cp config build_dir/.config COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=x86_64 SHELL=/bin/bash arch/x86/kernel/fpu/
If you fix the issue, kindly add following tag where applicable Reported-by: kernel test robot <lkp@intel.com>
All warnings (new ones prefixed by >>):
>> arch/x86/kernel/fpu/xstate.c:362:24: warning: format specifies type 'unsigned int' but the argument has type 'unsigned long' [-Wformat] "%u bytes\n", sizeof(union fpregs_state), init_fpstate.size); ^~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/printk.h:507:33: note: expanded from macro 'pr_err' printk(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__) ~~~ ^~~~~~~~~~~ include/linux/printk.h:464:60: note: expanded from macro 'printk' #define printk(fmt, ...) printk_index_wrap(_printk, fmt, ##__VA_ARGS__) ~~~ ^~~~~~~~~~~ include/linux/printk.h:436:19: note: expanded from macro 'printk_index_wrap' _p_func(_fmt, ##__VA_ARGS__); \ ~~~~ ^~~~~~~~~~~ 1 warning generated.
vim +362 arch/x86/kernel/fpu/xstate.c
328 329 /* 330 * All supported features have either init state all zeros or are 331 * handled in setup_init_fpu() individually. This is an explicit 332 * feature list and does not use XFEATURE_MASK*SUPPORTED to catch 333 * newly added supported features at build time and make people 334 * actually look at the init state for the new feature. 335 */ 336 #define XFEATURES_INIT_FPSTATE_HANDLED \ 337 (XFEATURE_MASK_FP | \ 338 XFEATURE_MASK_SSE | \ 339 XFEATURE_MASK_YMM | \ 340 XFEATURE_MASK_OPMASK | \ 341 XFEATURE_MASK_ZMM_Hi256 | \ 342 XFEATURE_MASK_Hi16_ZMM | \ 343 XFEATURE_MASK_PKRU | \ 344 XFEATURE_MASK_BNDREGS | \ 345 XFEATURE_MASK_BNDCSR | \ 346 XFEATURE_MASK_PASID | \ 347 XFEATURE_MASK_XTILE) 348 349 /* 350 * setup the xstate image representing the init state 351 */ 352 static int __init setup_init_fpu_buf(void) 353 { 354 BUILD_BUG_ON((XFEATURE_MASK_USER_SUPPORTED | 355 XFEATURE_MASK_SUPERVISOR_SUPPORTED) != 356 XFEATURES_INIT_FPSTATE_HANDLED); 357 358 print_xstate_features(); 359 360 if (init_fpstate.size > sizeof(union fpregs_state)) { 361 pr_err("x86/fpu: init_fpstate (%u bytes) is not enough for the configured size:" > 362 "%u bytes\n", sizeof(union fpregs_state), init_fpstate.size); 363 return -EINVAL; 364 } 365 366 xstate_init_xcomp_bv(&init_fpstate.regs.xsave, init_fpstate.xfeatures); 367 368 /* 369 * Init all the features state with header.xfeatures being 0x0 370 */ 371 os_xrstor_booting(&init_fpstate.regs.xsave); 372 373 /* 374 * All components are now in init state. Read the state back so 375 * that init_fpstate contains all non-zero init state. This only 376 * works with XSAVE, but not with XSAVEOPT and XSAVEC/S because 377 * those use the init optimization which skips writing data for 378 * components in init state. 379 * 380 * XSAVE could be used, but that would require to reshuffle the 381 * data when XSAVEC/S is available because XSAVEC/S uses xstate 382 * compaction. But doing so is a pointless exercise because most 383 * components have an all zeros init state except for the legacy 384 * ones (FP and SSE). Those can be saved with FXSAVE into the 385 * legacy area. Adding new features requires to ensure that init 386 * state is all zeroes or if not to add the necessary handling 387 * here. 388 */ 389 fxsave(&init_fpstate.regs.fxsave); 390 return 0; 391 } 392
-- 0-DAY CI Kernel Test Service https://01.org/lkp
| |