1# SPDX-License-Identifier: GPL-2.0 2menu "Kernel hacking" 3 4config TRACE_IRQFLAGS_SUPPORT 5 def_bool y 6 7source "lib/Kconfig.debug" 8 9config EARLY_PRINTK_USB 10 bool 11 12config X86_VERBOSE_BOOTUP 13 bool "Enable verbose x86 bootup info messages" 14 default y 15 ---help--- 16 Enables the informational output from the decompression stage 17 (e.g. bzImage) of the boot. If you disable this you will still 18 see errors. Disable this if you want silent bootup. 19 20config EARLY_PRINTK 21 bool "Early printk" if EXPERT 22 default y 23 ---help--- 24 Write kernel log output directly into the VGA buffer or to a serial 25 port. 26 27 This is useful for kernel debugging when your machine crashes very 28 early before the console code is initialized. For normal operation 29 it is not recommended because it looks ugly and doesn't cooperate 30 with klogd/syslogd or the X server. You should normally say N here, 31 unless you want to debug such a crash. 32 33config EARLY_PRINTK_DBGP 34 bool "Early printk via EHCI debug port" 35 depends on EARLY_PRINTK && PCI 36 select EARLY_PRINTK_USB 37 ---help--- 38 Write kernel log output directly into the EHCI debug port. 39 40 This is useful for kernel debugging when your machine crashes very 41 early before the console code is initialized. For normal operation 42 it is not recommended because it looks ugly and doesn't cooperate 43 with klogd/syslogd or the X server. You should normally say N here, 44 unless you want to debug such a crash. You need usb debug device. 45 46config EARLY_PRINTK_EFI 47 bool "Early printk via the EFI framebuffer" 48 depends on EFI && EARLY_PRINTK 49 select FONT_SUPPORT 50 ---help--- 51 Write kernel log output directly into the EFI framebuffer. 52 53 This is useful for kernel debugging when your machine crashes very 54 early before the console code is initialized. 55 56config EARLY_PRINTK_USB_XDBC 57 bool "Early printk via the xHCI debug port" 58 depends on EARLY_PRINTK && PCI 59 select EARLY_PRINTK_USB 60 ---help--- 61 Write kernel log output directly into the xHCI debug port. 62 63 One use for this feature is kernel debugging, for example when your 64 machine crashes very early before the regular console code is 65 initialized. Other uses include simpler, lockless logging instead of 66 a full-blown printk console driver + klogd. 67 68 For normal production environments this is normally not recommended, 69 because it doesn't feed events into klogd/syslogd and doesn't try to 70 print anything on the screen. 71 72 You should normally say N here, unless you want to debug early 73 crashes or need a very simple printk logging facility. 74 75config MCSAFE_TEST 76 def_bool n 77 78config X86_PTDUMP_CORE 79 def_bool n 80 81config X86_PTDUMP 82 tristate "Export kernel pagetable layout to userspace via debugfs" 83 depends on DEBUG_KERNEL 84 select DEBUG_FS 85 select X86_PTDUMP_CORE 86 ---help--- 87 Say Y here if you want to show the kernel pagetable layout in a 88 debugfs file. This information is only useful for kernel developers 89 who are working in architecture specific areas of the kernel. 90 It is probably not a good idea to enable this feature in a production 91 kernel. 92 If in doubt, say "N" 93 94config EFI_PGT_DUMP 95 bool "Dump the EFI pagetable" 96 depends on EFI 97 select X86_PTDUMP_CORE 98 ---help--- 99 Enable this if you want to dump the EFI page table before 100 enabling virtual mode. This can be used to debug miscellaneous 101 issues with the mapping of the EFI runtime regions into that 102 table. 103 104config DEBUG_WX 105 bool "Warn on W+X mappings at boot" 106 select X86_PTDUMP_CORE 107 ---help--- 108 Generate a warning if any W+X mappings are found at boot. 109 110 This is useful for discovering cases where the kernel is leaving 111 W+X mappings after applying NX, as such mappings are a security risk. 112 113 Look for a message in dmesg output like this: 114 115 x86/mm: Checked W+X mappings: passed, no W+X pages found. 116 117 or like this, if the check failed: 118 119 x86/mm: Checked W+X mappings: FAILED, <N> W+X pages found. 120 121 Note that even if the check fails, your kernel is possibly 122 still fine, as W+X mappings are not a security hole in 123 themselves, what they do is that they make the exploitation 124 of other unfixed kernel bugs easier. 125 126 There is no runtime or memory usage effect of this option 127 once the kernel has booted up - it's a one time check. 128 129 If in doubt, say "Y". 130 131config DOUBLEFAULT 132 default y 133 bool "Enable doublefault exception handler" if EXPERT 134 ---help--- 135 This option allows trapping of rare doublefault exceptions that 136 would otherwise cause a system to silently reboot. Disabling this 137 option saves about 4k and might cause you much additional grey 138 hair. 139 140config DEBUG_TLBFLUSH 141 bool "Set upper limit of TLB entries to flush one-by-one" 142 depends on DEBUG_KERNEL 143 ---help--- 144 145 X86-only for now. 146 147 This option allows the user to tune the amount of TLB entries the 148 kernel flushes one-by-one instead of doing a full TLB flush. In 149 certain situations, the former is cheaper. This is controlled by the 150 tlb_flushall_shift knob under /sys/kernel/debug/x86. If you set it 151 to -1, the code flushes the whole TLB unconditionally. Otherwise, 152 for positive values of it, the kernel will use single TLB entry 153 invalidating instructions according to the following formula: 154 155 flush_entries <= active_tlb_entries / 2^tlb_flushall_shift 156 157 If in doubt, say "N". 158 159config IOMMU_DEBUG 160 bool "Enable IOMMU debugging" 161 depends on GART_IOMMU && DEBUG_KERNEL 162 depends on X86_64 163 ---help--- 164 Force the IOMMU to on even when you have less than 4GB of 165 memory and add debugging code. On overflow always panic. And 166 allow to enable IOMMU leak tracing. Can be disabled at boot 167 time with iommu=noforce. This will also enable scatter gather 168 list merging. Currently not recommended for production 169 code. When you use it make sure you have a big enough 170 IOMMU/AGP aperture. Most of the options enabled by this can 171 be set more finegrained using the iommu= command line 172 options. See Documentation/x86/x86_64/boot-options.txt for more 173 details. 174 175config IOMMU_LEAK 176 bool "IOMMU leak tracing" 177 depends on IOMMU_DEBUG && DMA_API_DEBUG 178 ---help--- 179 Add a simple leak tracer to the IOMMU code. This is useful when you 180 are debugging a buggy device driver that leaks IOMMU mappings. 181 182config HAVE_MMIOTRACE_SUPPORT 183 def_bool y 184 185config X86_DECODER_SELFTEST 186 bool "x86 instruction decoder selftest" 187 depends on DEBUG_KERNEL && KPROBES 188 depends on !COMPILE_TEST 189 ---help--- 190 Perform x86 instruction decoder selftests at build time. 191 This option is useful for checking the sanity of x86 instruction 192 decoder code. 193 If unsure, say "N". 194 195# 196# IO delay types: 197# 198 199config IO_DELAY_TYPE_0X80 200 int 201 default "0" 202 203config IO_DELAY_TYPE_0XED 204 int 205 default "1" 206 207config IO_DELAY_TYPE_UDELAY 208 int 209 default "2" 210 211config IO_DELAY_TYPE_NONE 212 int 213 default "3" 214 215choice 216 prompt "IO delay type" 217 default IO_DELAY_0X80 218 219config IO_DELAY_0X80 220 bool "port 0x80 based port-IO delay [recommended]" 221 ---help--- 222 This is the traditional Linux IO delay used for in/out_p. 223 It is the most tested hence safest selection here. 224 225config IO_DELAY_0XED 226 bool "port 0xed based port-IO delay" 227 ---help--- 228 Use port 0xed as the IO delay. This frees up port 0x80 which is 229 often used as a hardware-debug port. 230 231config IO_DELAY_UDELAY 232 bool "udelay based port-IO delay" 233 ---help--- 234 Use udelay(2) as the IO delay method. This provides the delay 235 while not having any side-effect on the IO port space. 236 237config IO_DELAY_NONE 238 bool "no port-IO delay" 239 ---help--- 240 No port-IO delay. Will break on old boxes that require port-IO 241 delay for certain operations. Should work on most new machines. 242 243endchoice 244 245if IO_DELAY_0X80 246config DEFAULT_IO_DELAY_TYPE 247 int 248 default IO_DELAY_TYPE_0X80 249endif 250 251if IO_DELAY_0XED 252config DEFAULT_IO_DELAY_TYPE 253 int 254 default IO_DELAY_TYPE_0XED 255endif 256 257if IO_DELAY_UDELAY 258config DEFAULT_IO_DELAY_TYPE 259 int 260 default IO_DELAY_TYPE_UDELAY 261endif 262 263if IO_DELAY_NONE 264config DEFAULT_IO_DELAY_TYPE 265 int 266 default IO_DELAY_TYPE_NONE 267endif 268 269config DEBUG_BOOT_PARAMS 270 bool "Debug boot parameters" 271 depends on DEBUG_KERNEL 272 depends on DEBUG_FS 273 ---help--- 274 This option will cause struct boot_params to be exported via debugfs. 275 276config CPA_DEBUG 277 bool "CPA self-test code" 278 depends on DEBUG_KERNEL 279 ---help--- 280 Do change_page_attr() self-tests every 30 seconds. 281 282config OPTIMIZE_INLINING 283 bool "Allow gcc to uninline functions marked 'inline'" 284 ---help--- 285 This option determines if the kernel forces gcc to inline the functions 286 developers have marked 'inline'. Doing so takes away freedom from gcc to 287 do what it thinks is best, which is desirable for the gcc 3.x series of 288 compilers. The gcc 4.x series have a rewritten inlining algorithm and 289 enabling this option will generate a smaller kernel there. Hopefully 290 this algorithm is so good that allowing gcc 4.x and above to make the 291 decision will become the default in the future. Until then this option 292 is there to test gcc for this. 293 294 If unsure, say N. 295 296config DEBUG_ENTRY 297 bool "Debug low-level entry code" 298 depends on DEBUG_KERNEL 299 ---help--- 300 This option enables sanity checks in x86's low-level entry code. 301 Some of these sanity checks may slow down kernel entries and 302 exits or otherwise impact performance. 303 304 If unsure, say N. 305 306config DEBUG_NMI_SELFTEST 307 bool "NMI Selftest" 308 depends on DEBUG_KERNEL && X86_LOCAL_APIC 309 ---help--- 310 Enabling this option turns on a quick NMI selftest to verify 311 that the NMI behaves correctly. 312 313 This might help diagnose strange hangs that rely on NMI to 314 function properly. 315 316 If unsure, say N. 317 318config DEBUG_IMR_SELFTEST 319 bool "Isolated Memory Region self test" 320 default n 321 depends on INTEL_IMR 322 ---help--- 323 This option enables automated sanity testing of the IMR code. 324 Some simple tests are run to verify IMR bounds checking, alignment 325 and overlapping. This option is really only useful if you are 326 debugging an IMR memory map or are modifying the IMR code and want to 327 test your changes. 328 329 If unsure say N here. 330 331config X86_DEBUG_FPU 332 bool "Debug the x86 FPU code" 333 depends on DEBUG_KERNEL 334 default y 335 ---help--- 336 If this option is enabled then there will be extra sanity 337 checks and (boot time) debug printouts added to the kernel. 338 This debugging adds some small amount of runtime overhead 339 to the kernel. 340 341 If unsure, say N. 342 343config PUNIT_ATOM_DEBUG 344 tristate "ATOM Punit debug driver" 345 depends on PCI 346 select DEBUG_FS 347 select IOSF_MBI 348 ---help--- 349 This is a debug driver, which gets the power states 350 of all Punit North Complex devices. The power states of 351 each device is exposed as part of the debugfs interface. 352 The current power state can be read from 353 /sys/kernel/debug/punit_atom/dev_power_state 354 355choice 356 prompt "Choose kernel unwinder" 357 default UNWINDER_ORC if X86_64 358 default UNWINDER_FRAME_POINTER if X86_32 359 ---help--- 360 This determines which method will be used for unwinding kernel stack 361 traces for panics, oopses, bugs, warnings, perf, /proc/<pid>/stack, 362 livepatch, lockdep, and more. 363 364config UNWINDER_ORC 365 bool "ORC unwinder" 366 depends on X86_64 367 select STACK_VALIDATION 368 ---help--- 369 This option enables the ORC (Oops Rewind Capability) unwinder for 370 unwinding kernel stack traces. It uses a custom data format which is 371 a simplified version of the DWARF Call Frame Information standard. 372 373 This unwinder is more accurate across interrupt entry frames than the 374 frame pointer unwinder. It also enables a 5-10% performance 375 improvement across the entire kernel compared to frame pointers. 376 377 Enabling this option will increase the kernel's runtime memory usage 378 by roughly 2-4MB, depending on your kernel config. 379 380config UNWINDER_FRAME_POINTER 381 bool "Frame pointer unwinder" 382 select FRAME_POINTER 383 ---help--- 384 This option enables the frame pointer unwinder for unwinding kernel 385 stack traces. 386 387 The unwinder itself is fast and it uses less RAM than the ORC 388 unwinder, but the kernel text size will grow by ~3% and the kernel's 389 overall performance will degrade by roughly 5-10%. 390 391 This option is recommended if you want to use the livepatch 392 consistency model, as this is currently the only way to get a 393 reliable stack trace (CONFIG_HAVE_RELIABLE_STACKTRACE). 394 395config UNWINDER_GUESS 396 bool "Guess unwinder" 397 depends on EXPERT 398 depends on !STACKDEPOT 399 ---help--- 400 This option enables the "guess" unwinder for unwinding kernel stack 401 traces. It scans the stack and reports every kernel text address it 402 finds. Some of the addresses it reports may be incorrect. 403 404 While this option often produces false positives, it can still be 405 useful in many cases. Unlike the other unwinders, it has no runtime 406 overhead. 407 408endchoice 409 410config FRAME_POINTER 411 depends on !UNWINDER_ORC && !UNWINDER_GUESS 412 bool 413 414endmenu 415