1 # SPDX-License-Identifier: GPL-2.0
4 select ARCH_32BIT_OFF_T
5 select ARCH_HAS_BINFMT_FLAT if !MMU
6 select ARCH_HAS_DMA_PREP_COHERENT if MMU
7 select ARCH_HAS_SYNC_DMA_FOR_CPU if MMU
8 select ARCH_HAS_SYNC_DMA_FOR_DEVICE if MMU
9 select ARCH_HAS_DMA_SET_UNCACHED if MMU
10 select ARCH_USE_QUEUED_RWLOCKS
11 select ARCH_USE_QUEUED_SPINLOCKS
12 select ARCH_WANT_FRAME_POINTERS
13 select ARCH_WANT_IPC_PARSE_VERSION
14 select BUILDTIME_TABLE_SORT
15 select CLONE_BACKWARDS
17 select DMA_REMAP if MMU
18 select GENERIC_ATOMIC64
19 select GENERIC_IRQ_SHOW
20 select GENERIC_PCI_IOMAP
21 select GENERIC_SCHED_CLOCK
22 select GENERIC_STRNCPY_FROM_USER if KASAN
23 select HAVE_ARCH_AUDITSYSCALL
24 select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL
25 select HAVE_ARCH_KASAN if MMU && !XIP_KERNEL
26 select HAVE_ARCH_SECCOMP_FILTER
27 select HAVE_ARCH_TRACEHOOK
28 select HAVE_DEBUG_KMEMLEAK
29 select HAVE_DMA_CONTIGUOUS
30 select HAVE_EXIT_THREAD
31 select HAVE_FUNCTION_TRACER
32 select HAVE_FUTEX_CMPXCHG if !MMU
33 select HAVE_HW_BREAKPOINT if PERF_EVENTS
34 select HAVE_IRQ_TIME_ACCOUNTING
37 select HAVE_PERF_EVENTS
38 select HAVE_STACKPROTECTOR
39 select HAVE_SYSCALL_TRACEPOINTS
41 select MODULES_USE_ELF_RELA
42 select PERF_USE_VMALLOC
46 Xtensa processors are 32-bit RISC machines designed by Tensilica
47 primarily for embedded systems. These processors are both
48 configurable and extensible. The Linux port to the Xtensa
49 architecture supports all processor configurations and extensions,
50 with reasonable minimum requirements. The Xtensa Linux project has
51 a home page at <http://www.linux-xtensa.org/>.
53 config GENERIC_HWEIGHT
56 config ARCH_HAS_ILOG2_U32
59 config ARCH_HAS_ILOG2_U64
69 config LOCKDEP_SUPPORT
72 config STACKTRACE_SUPPORT
75 config TRACE_IRQFLAGS_SUPPORT
81 config HAVE_XTENSA_GPIO32
84 config KASAN_SHADOW_OFFSET
88 menu "Processor type and features"
91 prompt "Xtensa Processor Configuration"
92 default XTENSA_VARIANT_FSF
94 config XTENSA_VARIANT_FSF
95 bool "fsf - default (not generic) configuration"
98 config XTENSA_VARIANT_DC232B
99 bool "dc232b - Diamond 232L Standard Core Rev.B (LE)"
101 select HAVE_XTENSA_GPIO32
103 This variant refers to Tensilica's Diamond 232L Standard core Rev.B (LE).
105 config XTENSA_VARIANT_DC233C
106 bool "dc233c - Diamond 233L Standard Core Rev.C (LE)"
108 select HAVE_XTENSA_GPIO32
110 This variant refers to Tensilica's Diamond 233L Standard core Rev.C (LE).
112 config XTENSA_VARIANT_CUSTOM
113 bool "Custom Xtensa processor configuration"
114 select HAVE_XTENSA_GPIO32
116 Select this variant to use a custom Xtensa processor configuration.
117 You will be prompted for a processor variant CORENAME.
120 config XTENSA_VARIANT_CUSTOM_NAME
121 string "Xtensa Processor Custom Core Variant Name"
122 depends on XTENSA_VARIANT_CUSTOM
124 Provide the name of a custom Xtensa processor variant.
125 This CORENAME selects arch/xtensa/variant/CORENAME.
126 Don't forget you have to select MMU if you have one.
128 config XTENSA_VARIANT_NAME
130 default "dc232b" if XTENSA_VARIANT_DC232B
131 default "dc233c" if XTENSA_VARIANT_DC233C
132 default "fsf" if XTENSA_VARIANT_FSF
133 default XTENSA_VARIANT_CUSTOM_NAME if XTENSA_VARIANT_CUSTOM
135 config XTENSA_VARIANT_MMU
136 bool "Core variant has a Full MMU (TLB, Pages, Protection, etc)"
137 depends on XTENSA_VARIANT_CUSTOM
141 Build a Conventional Kernel with full MMU support,
142 ie: it supports a TLB with auto-loading, page protection.
144 config XTENSA_VARIANT_HAVE_PERF_EVENTS
145 bool "Core variant has Performance Monitor Module"
146 depends on XTENSA_VARIANT_CUSTOM
149 Enable if core variant has Performance Monitor Module with
150 External Registers Interface.
154 config XTENSA_FAKE_NMI
155 bool "Treat PMM IRQ as NMI"
156 depends on XTENSA_VARIANT_HAVE_PERF_EVENTS
159 If PMM IRQ is the only IRQ at EXCM level it is safe to
160 treat it as NMI, which improves accuracy of profiling.
162 If there are other interrupts at or above PMM IRQ priority level
163 but not above the EXCM level, PMM IRQ still may be treated as NMI,
164 but only if these IRQs are not used. There will be a build warning
165 saying that this is not safe, and a bugcheck if one of these IRQs
170 config XTENSA_UNALIGNED_USER
171 bool "Unaligned memory access in user space"
173 The Xtensa architecture currently does not handle unaligned
174 memory accesses in hardware but through an exception handler.
175 Per default, unaligned memory accesses are disabled in user space.
177 Say Y here to enable unaligned memory access in user space.
180 bool "System Supports SMP (MX)"
181 depends on XTENSA_VARIANT_CUSTOM
184 This option is used to indicate that the system-on-a-chip (SOC)
185 supports Multiprocessing. Multiprocessor support implemented above
186 the CPU core definition and currently needs to be selected manually.
188 Multiprocessor support is implemented with external cache and
189 interrupt controllers.
191 The MX interrupt distributer adds Interprocessor Interrupts
192 and causes the IRQ numbers to be increased by 4 for devices
193 like the open cores ethernet driver and the serial interface.
195 You still have to select "Enable SMP" to enable SMP on this SOC.
198 bool "Enable Symmetric multi-processing support"
200 select GENERIC_SMP_IDLE_THREAD
202 Enabled SMP Software; allows more than one CPU/CORE
203 to be activated during startup.
207 int "Maximum number of CPUs (2-32)"
212 bool "Enable CPU hotplug support"
215 Say Y here to allow turning CPUs off and on. CPUs can be
216 controlled through /sys/devices/system/cpu.
218 Say N if you want to disable CPU hotplug.
220 config FAST_SYSCALL_XTENSA
221 bool "Enable fast atomic syscalls"
224 fast_syscall_xtensa is a syscall that can make atomic operations
225 on UP kernel when processor has no s32c1i support.
227 This syscall is deprecated. It may have issues when called with
228 invalid arguments. It is provided only for backwards compatibility.
229 Only enable it if your userspace software requires it.
233 config FAST_SYSCALL_SPILL_REGISTERS
234 bool "Enable spill registers syscall"
237 fast_syscall_spill_registers is a syscall that spills all active
238 register windows of a calling userspace task onto its stack.
240 This syscall is deprecated. It may have issues when called with
241 invalid arguments. It is provided only for backwards compatibility.
242 Only enable it if your userspace software requires it.
246 config USER_ABI_CALL0
250 prompt "Userspace ABI"
251 default USER_ABI_DEFAULT
253 Select supported userspace ABI.
255 If unsure, choose the default ABI.
257 config USER_ABI_DEFAULT
258 bool "Default ABI only"
260 Assume default userspace ABI. For XEA2 cores it is windowed ABI.
261 call0 ABI binaries may be run on such kernel, but signal delivery
262 will not work correctly for them.
264 config USER_ABI_CALL0_ONLY
265 bool "Call0 ABI only"
266 select USER_ABI_CALL0
268 Select this option to support only call0 ABI in userspace.
269 Windowed ABI binaries will crash with a segfault caused by
270 an illegal instruction exception on the first 'entry' opcode.
272 Choose this option if you're planning to run only user code
273 built with call0 ABI.
275 config USER_ABI_CALL0_PROBE
276 bool "Support both windowed and call0 ABI by probing"
277 select USER_ABI_CALL0
279 Select this option to support both windowed and call0 userspace
280 ABIs. When enabled all processes are started with PS.WOE disabled
281 and a fast user exception handler for an illegal instruction is
282 used to turn on PS.WOE bit on the first 'entry' opcode executed by
285 This option should be enabled for the kernel that must support
286 both call0 and windowed ABIs in userspace at the same time.
288 Note that Xtensa ISA does not guarantee that entry opcode will
289 raise an illegal instruction exception on cores with XEA2 when
290 PS.WOE is disabled, check whether the target core supports it.
296 config XTENSA_CALIBRATE_CCOUNT
299 On some platforms (XT2000, for example), the CPU clock rate can
300 vary. The frequency can be determined, however, by measuring
301 against a well known, fixed frequency, such as an UART oscillator.
303 config SERIAL_CONSOLE
306 config PLATFORM_HAVE_XIP
309 menu "Platform options"
312 prompt "Xtensa System Type"
313 default XTENSA_PLATFORM_ISS
315 config XTENSA_PLATFORM_ISS
317 select XTENSA_CALIBRATE_CCOUNT
318 select SERIAL_CONSOLE
320 ISS is an acronym for Tensilica's Instruction Set Simulator.
322 config XTENSA_PLATFORM_XT2000
326 XT2000 is the name of Tensilica's feature-rich emulation platform.
327 This hardware is capable of running a full Linux distribution.
329 config XTENSA_PLATFORM_XTFPGA
331 select ETHOC if ETHERNET
332 select PLATFORM_WANT_DEFAULT_MEM if !MMU
333 select SERIAL_CONSOLE
334 select XTENSA_CALIBRATE_CCOUNT
335 select PLATFORM_HAVE_XIP
337 XTFPGA is the name of Tensilica board family (LX60, LX110, LX200, ML605).
338 This hardware is capable of running a full Linux distribution.
342 config PLATFORM_NR_IRQS
344 default 3 if XTENSA_PLATFORM_XT2000
347 config XTENSA_CPU_CLOCK
348 int "CPU clock rate [MHz]"
349 depends on !XTENSA_CALIBRATE_CCOUNT
352 config GENERIC_CALIBRATE_DELAY
353 bool "Auto calibration of the BogoMIPS value"
355 The BogoMIPS value can easily be derived from the CPU frequency.
358 bool "Default bootloader kernel arguments"
361 string "Initial kernel command string"
362 depends on CMDLINE_BOOL
363 default "console=ttyS0,38400 root=/dev/ram"
365 On some architectures (EBSA110 and CATS), there is currently no way
366 for the boot loader to pass arguments to the kernel. For these
367 architectures, you should supply some command-line options at build
368 time by entering them here. As a minimum, you should specify the
369 memory size and the root device (e.g., mem=64M root=/dev/nfs).
372 bool "Flattened Device Tree support"
374 select OF_EARLY_FLATTREE
376 Include support for flattened device tree machine descriptions.
378 config BUILTIN_DTB_SOURCE
379 string "DTB to build into the kernel image"
382 config PARSE_BOOTPARAM
383 bool "Parse bootparam block"
386 Parse parameters passed to the kernel from the bootloader. It may
387 be disabled if the kernel is known to run without the bootloader.
391 config BLK_DEV_SIMDISK
392 tristate "Host file-based simulated block device support"
394 depends on XTENSA_PLATFORM_ISS && BLOCK
396 Create block devices that map to files in the host file system.
397 Device binding to host file may be changed at runtime via proc
398 interface provided the device is not in use.
400 config BLK_DEV_SIMDISK_COUNT
401 int "Number of host file-based simulated block devices"
403 depends on BLK_DEV_SIMDISK
406 This is the default minimal number of created block devices.
407 Kernel/module parameter 'simdisk_count' may be used to change this
408 value at runtime. More file names (but no more than 10) may be
409 specified as parameters, simdisk_count grows accordingly.
411 config SIMDISK0_FILENAME
412 string "Host filename for the first simulated device"
413 depends on BLK_DEV_SIMDISK = y
416 Attach a first simdisk to a host file. Conventionally, this file
417 contains a root file system.
419 config SIMDISK1_FILENAME
420 string "Host filename for the second simulated device"
421 depends on BLK_DEV_SIMDISK = y && BLK_DEV_SIMDISK_COUNT != 1
424 Another simulated disk in a host file for a buildroot-independent
428 bool "Enable XTFPGA LCD driver"
429 depends on XTENSA_PLATFORM_XTFPGA
432 There's a 2x16 LCD on most of XTFPGA boards, kernel may output
433 progress messages there during bootup/shutdown. It may be useful
434 during board bringup.
438 config XTFPGA_LCD_BASE_ADDR
439 hex "XTFPGA LCD base address"
440 depends on XTFPGA_LCD
443 Base address of the LCD controller inside KIO region.
444 Different boards from XTFPGA family have LCD controller at different
445 addresses. Please consult prototyping user guide for your board for
446 the correct address. Wrong address here may lead to hardware lockup.
448 config XTFPGA_LCD_8BIT_ACCESS
449 bool "Use 8-bit access to XTFPGA LCD"
450 depends on XTFPGA_LCD
453 LCD may be connected with 4- or 8-bit interface, 8-bit access may
454 only be used with 8-bit interface. Please consult prototyping user
455 guide for your board for the correct interface width.
457 comment "Kernel memory layout"
459 config INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
460 bool "Initialize Xtensa MMU inside the Linux kernel code"
461 depends on !XTENSA_VARIANT_FSF && !XTENSA_VARIANT_DC232B
462 default y if XTENSA_VARIANT_DC233C || XTENSA_VARIANT_CUSTOM
464 Earlier version initialized the MMU in the exception vector
465 before jumping to _startup in head.S and had an advantage that
466 it was possible to place a software breakpoint at 'reset' and
467 then enter your normal kernel breakpoints once the MMU was mapped
468 to the kernel mappings (0XC0000000).
470 This unfortunately won't work for U-Boot and likely also wont
471 work for using KEXEC to have a hot kernel ready for doing a
474 So now the MMU is initialized in head.S but it's necessary to
475 use hardware breakpoints (gdb 'hbreak' cmd) to break at _startup.
476 xt-gdb can't place a Software Breakpoint in the 0XD region prior
477 to mapping the MMU and after mapping even if the area of low memory
478 was mapped gdb wouldn't remove the breakpoint on hitting it as the
479 PC wouldn't match. Since Hardware Breakpoints are recommended for
480 Linux configurations it seems reasonable to just assume they exist
481 and leave this older mechanism for unfortunate souls that choose
482 not to follow Tensilica's recommendation.
484 Selecting this will cause U-Boot to set the KERNEL Load and Entry
485 address at 0x00003000 instead of the mapped std of 0xD0003000.
490 bool "Kernel Execute-In-Place from ROM"
491 depends on PLATFORM_HAVE_XIP
493 Execute-In-Place allows the kernel to run from non-volatile storage
494 directly addressable by the CPU, such as NOR flash. This saves RAM
495 space since the text section of the kernel is not loaded from flash
496 to RAM. Read-write sections, such as the data section and stack,
497 are still copied to RAM. The XIP kernel is not compressed since
498 it has to run directly from flash, so it will take more space to
499 store it. The flash address used to link the kernel object files,
500 and for storing it, is configuration dependent. Therefore, if you
501 say Y here, you must know the proper physical address where to
502 store the kernel image depending on your own flash memory usage.
504 Also note that the make target becomes "make xipImage" rather than
505 "make Image" or "make uImage". The final kernel binary to put in
506 ROM memory will be arch/xtensa/boot/xipImage.
510 config MEMMAP_CACHEATTR
511 hex "Cache attributes for the memory address space"
515 These cache attributes are set up for noMMU systems. Each hex digit
516 specifies cache attributes for the corresponding 512MB memory
517 region: bits 0..3 -- for addresses 0x00000000..0x1fffffff,
518 bits 4..7 -- for addresses 0x20000000..0x3fffffff, and so on.
520 Cache attribute values are specific for the MMU type.
521 For region protection MMUs:
533 3: special (c and e are illegal, f is reserved).
537 2: WB, no-write-allocate cache,
542 hex "Physical address of the KSEG mapping"
543 depends on INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX && MMU
546 This is the physical address where KSEG is mapped. Please refer to
547 the chosen KSEG layout help for the required address alignment.
548 Unpacked kernel image (including vectors) must be located completely
550 Physical memory below this address is not available to linux.
552 If unsure, leave the default value here.
554 config KERNEL_VIRTUAL_ADDRESS
555 hex "Kernel virtual address"
556 depends on MMU && XIP_KERNEL
559 This is the virtual address where the XIP kernel is mapped.
560 XIP kernel may be mapped into KSEG or KIO region, virtual address
561 provided here must match kernel load address provided in
564 config KERNEL_LOAD_ADDRESS
565 hex "Kernel load address"
566 default 0x60003000 if !MMU
567 default 0x00003000 if MMU && INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
568 default 0xd0003000 if MMU && !INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
570 This is the address where the kernel is loaded.
571 It is virtual address for MMUv2 configurations and physical address
572 for all other configurations.
574 If unsure, leave the default value here.
577 prompt "Relocatable vectors location"
578 default XTENSA_VECTORS_IN_TEXT
580 Choose whether relocatable vectors are merged into the kernel .text
581 or placed separately at runtime. This option does not affect
582 configurations without VECBASE register where vectors are always
583 placed at their hardware-defined locations.
585 config XTENSA_VECTORS_IN_TEXT
586 bool "Merge relocatable vectors into kernel text"
589 This option puts relocatable vectors into the kernel .text section
590 with proper alignment.
591 This is a safe choice for most configurations.
593 config XTENSA_VECTORS_SEPARATE
594 bool "Put relocatable vectors at fixed address"
596 This option puts relocatable vectors at specific virtual address.
597 Vectors are merged with the .init data in the kernel image and
598 are copied into their designated location during kernel startup.
599 Use it to put vectors into IRAM or out of FLASH on kernels with
600 XIP-aware MTD support.
605 hex "Kernel vectors virtual address"
607 depends on XTENSA_VECTORS_SEPARATE
609 This is the virtual address of the (relocatable) vectors base.
610 It must be within KSEG if MMU is used.
613 hex "XIP kernel data virtual address"
614 depends on XIP_KERNEL
617 This is the virtual address where XIP kernel data is copied.
618 It must be within KSEG if MMU is used.
620 config PLATFORM_WANT_DEFAULT_MEM
623 config DEFAULT_MEM_START
625 prompt "PAGE_OFFSET/PHYS_OFFSET" if !MMU && PLATFORM_WANT_DEFAULT_MEM
626 default 0x60000000 if PLATFORM_WANT_DEFAULT_MEM
629 This is the base address used for both PAGE_OFFSET and PHYS_OFFSET
630 in noMMU configurations.
632 If unsure, leave the default value here.
637 default XTENSA_KSEG_MMU_V2
639 config XTENSA_KSEG_MMU_V2
640 bool "MMUv2: 128MB cached + 128MB uncached"
642 MMUv2 compatible kernel memory map: TLB way 5 maps 128MB starting
643 at KSEG_PADDR to 0xd0000000 with cache and to 0xd8000000
645 KSEG_PADDR must be aligned to 128MB.
647 config XTENSA_KSEG_256M
648 bool "256MB cached + 256MB uncached"
649 depends on INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
651 TLB way 6 maps 256MB starting at KSEG_PADDR to 0xb0000000
652 with cache and to 0xc0000000 without cache.
653 KSEG_PADDR must be aligned to 256MB.
655 config XTENSA_KSEG_512M
656 bool "512MB cached + 512MB uncached"
657 depends on INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
659 TLB way 6 maps 512MB starting at KSEG_PADDR to 0xa0000000
660 with cache and to 0xc0000000 without cache.
661 KSEG_PADDR must be aligned to 256MB.
666 bool "High Memory Support"
670 Linux can use the full amount of RAM in the system by
671 default. However, the default MMUv2 setup only maps the
672 lowermost 128 MB of memory linearly to the areas starting
673 at 0xd0000000 (cached) and 0xd8000000 (uncached).
674 When there are more than 128 MB memory in the system not
675 all of it can be "permanently mapped" by the kernel.
676 The physical memory that's not permanently mapped is called
679 If you are compiling a kernel which will never run on a
680 machine with more than 128 MB total physical RAM, answer
685 config FORCE_MAX_ZONEORDER
686 int "Maximum zone order"
689 The kernel memory allocator divides physically contiguous memory
690 blocks into "zones", where each zone is a power of two number of
691 pages. This option selects the largest power of two that the kernel
692 keeps in the memory allocator. If you need to allocate very large
693 blocks of physically contiguous memory, then you may need to
696 This config option is actually maximum order plus one. For example,
697 a value of 11 means that the largest free memory block is 2^10 pages.
701 menu "Power management options"
703 source "kernel/power/Kconfig"