1menu "Kernel hacking" 2 3config TRACE_IRQFLAGS_SUPPORT 4 def_bool y 5 6source "lib/Kconfig.debug" 7 8config STRICT_DEVMEM 9 bool "Filter access to /dev/mem" 10 ---help--- 11 If this option is disabled, you allow userspace (root) access to all 12 of memory, including kernel and userspace memory. Accidental 13 access to this is obviously disastrous, but specific access can 14 be used by people debugging the kernel. Note that with PAT support 15 enabled, even in this case there are restrictions on /dev/mem 16 use due to the cache aliasing requirements. 17 18 If this option is switched on, the /dev/mem file only allows 19 userspace access to PCI space and the BIOS code and data regions. 20 This is sufficient for dosemu and X and all common users of 21 /dev/mem. 22 23 If in doubt, say Y. 24 25config X86_VERBOSE_BOOTUP 26 bool "Enable verbose x86 bootup info messages" 27 default y 28 ---help--- 29 Enables the informational output from the decompression stage 30 (e.g. bzImage) of the boot. If you disable this you will still 31 see errors. Disable this if you want silent bootup. 32 33config EARLY_PRINTK 34 bool "Early printk" if EXPERT 35 default y 36 ---help--- 37 Write kernel log output directly into the VGA buffer or to a serial 38 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 N here, 44 unless you want to debug such a crash. 45 46config EARLY_PRINTK_DBGP 47 bool "Early printk via EHCI debug port" 48 depends on EARLY_PRINTK && PCI 49 ---help--- 50 Write kernel log output directly into the EHCI debug port. 51 52 This is useful for kernel debugging when your machine crashes very 53 early before the console code is initialized. For normal operation 54 it is not recommended because it looks ugly and doesn't cooperate 55 with klogd/syslogd or the X server. You should normally N here, 56 unless you want to debug such a crash. You need usb debug device. 57 58config EARLY_PRINTK_EFI 59 bool "Early printk via the EFI framebuffer" 60 depends on EFI && EARLY_PRINTK 61 select FONT_SUPPORT 62 ---help--- 63 Write kernel log output directly into the EFI framebuffer. 64 65 This is useful for kernel debugging when your machine crashes very 66 early before the console code is initialized. 67 68config X86_PTDUMP_CORE 69 def_bool n 70 71config X86_PTDUMP 72 bool "Export kernel pagetable layout to userspace via debugfs" 73 depends on DEBUG_KERNEL 74 select DEBUG_FS 75 select X86_PTDUMP_CORE 76 ---help--- 77 Say Y here if you want to show the kernel pagetable layout in a 78 debugfs file. This information is only useful for kernel developers 79 who are working in architecture specific areas of the kernel. 80 It is probably not a good idea to enable this feature in a production 81 kernel. 82 If in doubt, say "N" 83 84config EFI_PGT_DUMP 85 bool "Dump the EFI pagetable" 86 depends on EFI 87 select X86_PTDUMP_CORE 88 ---help--- 89 Enable this if you want to dump the EFI page table before 90 enabling virtual mode. This can be used to debug miscellaneous 91 issues with the mapping of the EFI runtime regions into that 92 table. 93 94config DEBUG_RODATA 95 bool "Write protect kernel read-only data structures" 96 default y 97 depends on DEBUG_KERNEL 98 ---help--- 99 Mark the kernel read-only data as write-protected in the pagetables, 100 in order to catch accidental (and incorrect) writes to such const 101 data. This is recommended so that we can catch kernel bugs sooner. 102 If in doubt, say "Y". 103 104config DEBUG_RODATA_TEST 105 bool "Testcase for the DEBUG_RODATA feature" 106 depends on DEBUG_RODATA 107 default y 108 ---help--- 109 This option enables a testcase for the DEBUG_RODATA 110 feature as well as for the change_page_attr() infrastructure. 111 If in doubt, say "N" 112 113config DEBUG_WX 114 bool "Warn on W+X mappings at boot" 115 depends on DEBUG_RODATA 116 select X86_PTDUMP_CORE 117 ---help--- 118 Generate a warning if any W+X mappings are found at boot. 119 120 This is useful for discovering cases where the kernel is leaving 121 W+X mappings after applying NX, as such mappings are a security risk. 122 123 Look for a message in dmesg output like this: 124 125 x86/mm: Checked W+X mappings: passed, no W+X pages found. 126 127 or like this, if the check failed: 128 129 x86/mm: Checked W+X mappings: FAILED, <N> W+X pages found. 130 131 Note that even if the check fails, your kernel is possibly 132 still fine, as W+X mappings are not a security hole in 133 themselves, what they do is that they make the exploitation 134 of other unfixed kernel bugs easier. 135 136 There is no runtime or memory usage effect of this option 137 once the kernel has booted up - it's a one time check. 138 139 If in doubt, say "Y". 140 141config DEBUG_SET_MODULE_RONX 142 bool "Set loadable kernel module data as NX and text as RO" 143 depends on MODULES 144 ---help--- 145 This option helps catch unintended modifications to loadable 146 kernel module's text and read-only data. It also prevents execution 147 of module data. Such protection may interfere with run-time code 148 patching and dynamic kernel tracing - and they might also protect 149 against certain classes of kernel exploits. 150 If in doubt, say "N". 151 152config DEBUG_NX_TEST 153 tristate "Testcase for the NX non-executable stack feature" 154 depends on DEBUG_KERNEL && m 155 ---help--- 156 This option enables a testcase for the CPU NX capability 157 and the software setup of this feature. 158 If in doubt, say "N" 159 160config DOUBLEFAULT 161 default y 162 bool "Enable doublefault exception handler" if EXPERT 163 ---help--- 164 This option allows trapping of rare doublefault exceptions that 165 would otherwise cause a system to silently reboot. Disabling this 166 option saves about 4k and might cause you much additional grey 167 hair. 168 169config DEBUG_TLBFLUSH 170 bool "Set upper limit of TLB entries to flush one-by-one" 171 depends on DEBUG_KERNEL 172 ---help--- 173 174 X86-only for now. 175 176 This option allows the user to tune the amount of TLB entries the 177 kernel flushes one-by-one instead of doing a full TLB flush. In 178 certain situations, the former is cheaper. This is controlled by the 179 tlb_flushall_shift knob under /sys/kernel/debug/x86. If you set it 180 to -1, the code flushes the whole TLB unconditionally. Otherwise, 181 for positive values of it, the kernel will use single TLB entry 182 invalidating instructions according to the following formula: 183 184 flush_entries <= active_tlb_entries / 2^tlb_flushall_shift 185 186 If in doubt, say "N". 187 188config IOMMU_DEBUG 189 bool "Enable IOMMU debugging" 190 depends on GART_IOMMU && DEBUG_KERNEL 191 depends on X86_64 192 ---help--- 193 Force the IOMMU to on even when you have less than 4GB of 194 memory and add debugging code. On overflow always panic. And 195 allow to enable IOMMU leak tracing. Can be disabled at boot 196 time with iommu=noforce. This will also enable scatter gather 197 list merging. Currently not recommended for production 198 code. When you use it make sure you have a big enough 199 IOMMU/AGP aperture. Most of the options enabled by this can 200 be set more finegrained using the iommu= command line 201 options. See Documentation/x86/x86_64/boot-options.txt for more 202 details. 203 204config IOMMU_STRESS 205 bool "Enable IOMMU stress-test mode" 206 ---help--- 207 This option disables various optimizations in IOMMU related 208 code to do real stress testing of the IOMMU code. This option 209 will cause a performance drop and should only be enabled for 210 testing. 211 212config IOMMU_LEAK 213 bool "IOMMU leak tracing" 214 depends on IOMMU_DEBUG && DMA_API_DEBUG 215 ---help--- 216 Add a simple leak tracer to the IOMMU code. This is useful when you 217 are debugging a buggy device driver that leaks IOMMU mappings. 218 219config HAVE_MMIOTRACE_SUPPORT 220 def_bool y 221 222config X86_DECODER_SELFTEST 223 bool "x86 instruction decoder selftest" 224 depends on DEBUG_KERNEL && KPROBES 225 depends on !COMPILE_TEST 226 ---help--- 227 Perform x86 instruction decoder selftests at build time. 228 This option is useful for checking the sanity of x86 instruction 229 decoder code. 230 If unsure, say "N". 231 232# 233# IO delay types: 234# 235 236config IO_DELAY_TYPE_0X80 237 int 238 default "0" 239 240config IO_DELAY_TYPE_0XED 241 int 242 default "1" 243 244config IO_DELAY_TYPE_UDELAY 245 int 246 default "2" 247 248config IO_DELAY_TYPE_NONE 249 int 250 default "3" 251 252choice 253 prompt "IO delay type" 254 default IO_DELAY_0X80 255 256config IO_DELAY_0X80 257 bool "port 0x80 based port-IO delay [recommended]" 258 ---help--- 259 This is the traditional Linux IO delay used for in/out_p. 260 It is the most tested hence safest selection here. 261 262config IO_DELAY_0XED 263 bool "port 0xed based port-IO delay" 264 ---help--- 265 Use port 0xed as the IO delay. This frees up port 0x80 which is 266 often used as a hardware-debug port. 267 268config IO_DELAY_UDELAY 269 bool "udelay based port-IO delay" 270 ---help--- 271 Use udelay(2) as the IO delay method. This provides the delay 272 while not having any side-effect on the IO port space. 273 274config IO_DELAY_NONE 275 bool "no port-IO delay" 276 ---help--- 277 No port-IO delay. Will break on old boxes that require port-IO 278 delay for certain operations. Should work on most new machines. 279 280endchoice 281 282if IO_DELAY_0X80 283config DEFAULT_IO_DELAY_TYPE 284 int 285 default IO_DELAY_TYPE_0X80 286endif 287 288if IO_DELAY_0XED 289config DEFAULT_IO_DELAY_TYPE 290 int 291 default IO_DELAY_TYPE_0XED 292endif 293 294if IO_DELAY_UDELAY 295config DEFAULT_IO_DELAY_TYPE 296 int 297 default IO_DELAY_TYPE_UDELAY 298endif 299 300if IO_DELAY_NONE 301config DEFAULT_IO_DELAY_TYPE 302 int 303 default IO_DELAY_TYPE_NONE 304endif 305 306config DEBUG_BOOT_PARAMS 307 bool "Debug boot parameters" 308 depends on DEBUG_KERNEL 309 depends on DEBUG_FS 310 ---help--- 311 This option will cause struct boot_params to be exported via debugfs. 312 313config CPA_DEBUG 314 bool "CPA self-test code" 315 depends on DEBUG_KERNEL 316 ---help--- 317 Do change_page_attr() self-tests every 30 seconds. 318 319config OPTIMIZE_INLINING 320 bool "Allow gcc to uninline functions marked 'inline'" 321 ---help--- 322 This option determines if the kernel forces gcc to inline the functions 323 developers have marked 'inline'. Doing so takes away freedom from gcc to 324 do what it thinks is best, which is desirable for the gcc 3.x series of 325 compilers. The gcc 4.x series have a rewritten inlining algorithm and 326 enabling this option will generate a smaller kernel there. Hopefully 327 this algorithm is so good that allowing gcc 4.x and above to make the 328 decision will become the default in the future. Until then this option 329 is there to test gcc for this. 330 331 If unsure, say N. 332 333config DEBUG_ENTRY 334 bool "Debug low-level entry code" 335 depends on DEBUG_KERNEL 336 ---help--- 337 This option enables sanity checks in x86's low-level entry code. 338 Some of these sanity checks may slow down kernel entries and 339 exits or otherwise impact performance. 340 341 This is currently used to help test NMI code. 342 343 If unsure, say N. 344 345config DEBUG_NMI_SELFTEST 346 bool "NMI Selftest" 347 depends on DEBUG_KERNEL && X86_LOCAL_APIC 348 ---help--- 349 Enabling this option turns on a quick NMI selftest to verify 350 that the NMI behaves correctly. 351 352 This might help diagnose strange hangs that rely on NMI to 353 function properly. 354 355 If unsure, say N. 356 357config DEBUG_IMR_SELFTEST 358 bool "Isolated Memory Region self test" 359 default n 360 depends on INTEL_IMR 361 ---help--- 362 This option enables automated sanity testing of the IMR code. 363 Some simple tests are run to verify IMR bounds checking, alignment 364 and overlapping. This option is really only useful if you are 365 debugging an IMR memory map or are modifying the IMR code and want to 366 test your changes. 367 368 If unsure say N here. 369 370config X86_DEBUG_STATIC_CPU_HAS 371 bool "Debug alternatives" 372 depends on DEBUG_KERNEL 373 ---help--- 374 This option causes additional code to be generated which 375 fails if static_cpu_has() is used before alternatives have 376 run. 377 378 If unsure, say N. 379 380config X86_DEBUG_FPU 381 bool "Debug the x86 FPU code" 382 depends on DEBUG_KERNEL 383 default y 384 ---help--- 385 If this option is enabled then there will be extra sanity 386 checks and (boot time) debug printouts added to the kernel. 387 This debugging adds some small amount of runtime overhead 388 to the kernel. 389 390 If unsure, say N. 391 392config PUNIT_ATOM_DEBUG 393 tristate "ATOM Punit debug driver" 394 select DEBUG_FS 395 select IOSF_MBI 396 ---help--- 397 This is a debug driver, which gets the power states 398 of all Punit North Complex devices. The power states of 399 each device is exposed as part of the debugfs interface. 400 The current power state can be read from 401 /sys/kernel/debug/punit_atom/dev_power_state 402 403endmenu 404