- 09 Feb, 2021 5 commits
-
-
Serge Semin authored
This reverts commit 3751cbda. Originally the patch was created to fix the reserved-memory DT-node parsing failure on the early stages of the platform memory initialization. That happened due to the two early memory allocators utilization that time: bootmem and memblock. At first the platform-specific memory mapping array was initialized. Then the early_init_fdt_scan_reserved_mem() was called, which couldn't fully parse the "reserved-memory" DT-node since neither memblock nor bootmem allocators hadn't been initialized at that stage, so the fdt_init_reserved_mem() method failed on the memory allocation calls. Only after that the platform-specific memory mapping were used to create proper bootmem and memblock structures and let the early memory allocations work. That's why we had to call the fdt_init_reserved_mem() method one more time to retry the initialization of the features like CMA. The necessity to have that fix was disappeared after the full memblock support had been added to the MIPS kernel and all plat_mem_setup() had been fixed to add the memory regions right into the memblock memory pool. Let's revert that patch then especially after having Paul reported that the second fdt_init_reserved_mem() call causes the reserved memory pool being created twice bigger than implied. Fixes: a94e4f24 ("MIPS: init: Drop boot_mem_map") Reported-by:
Paul Cercueil <paul@crapouillou.net> Signed-off-by:
Serge Semin <Sergey.Semin@baikalelectronics.ru> Signed-off-by:
Thomas Bogendoerfer <tsbogend@alpha.franken.de>
-
Jinyang He authored
unwind_stack_by_address and unwind_stack need <asm/stacktrace.h>. arch_align_stack needs <asm/exec.h> link: https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org/thread/ZPL2RRA6RZKRQZI5IGOVLFXN2GVZBN3L/Reported-by:
kernel test robot <lkp@intel.com> Signed-off-by:
Jinyang He <hejinyang@loongson.cn> Signed-off-by:
-
Alexander Lobakin authored
Commit 1e35918a ("MIPS: Enable Undefined Behavior Sanitizer UBSAN") added a possibility to build the entire kernel with UBSAN instrumentation for MIPS, with the exception for VDSO. However, self-extracting head wasn't been added to exceptions, so this occurs: mips-alpine-linux-musl-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_buildDTable_wksp': decompress.c:(.text.FSE_buildDTable_wksp+0x278): undefined reference to `__ubsan_handle_shift_out_of_bounds' mips-alpine-linux-musl-ld: decompress.c:(.text.FSE_buildDTable_wksp+0x2a8): undefined reference to `__ubsan_handle_shift_out_of_bounds' mips-alpine-linux-musl-ld: decompress.c:(.text.FSE_buildDTable_wksp+0x2c4): undefined reference to `__ubsan_handle_shift_out_of_bounds' mips-alpine-linux-musl-ld: arch/mips/boot/compressed/decompress.o: decompress.c:(.text.FSE_buildDTable_raw+0x9c): more undefined references to `__ubsan_handle_shift_out_of_bounds' follow Add UBSAN_SANITIZE := n to mips/boot/compressed/Makefile to exclude it from instrumentation scope and fix this issue. Fixes: 1e35918a ("MIPS: Enable Undefined Behavior Sanitizer UBSAN") Cc: stable@vger.kernel.org # 5.0+ Signed-off-by:
Alexander Lobakin <alobakin@pm.me> Signed-off-by:
-
Jinyang He authored
The type of the VMLINUX_LOAD_ADDRESS macro is the (unsigned long long) in 32bits kernel but (unsigned long) in the 64-bit kernel. Although there is no error here, avoid using it to calculate kaslr_offset. Signed-off-by:
-
Jinyang He authored
Provide kaslr_offset() to get the kernel offset when KASLR is enabled. Error may occur before update_kaslr_offset(), so put it at the end of the offset branch. Fixes: a307a4ce ("MIPS: Loongson64: Add KASLR support") Reported-by:
-
- 04 Feb, 2021 9 commits
-
-
Tiezhu Yang authored
Add perf_event_mips_regs/perf_reg_value/perf_reg_validate to support features HAVE_PERF_REGS/HAVE_PERF_USER_STACK_DUMP in kernel. [ayan@wavecomp.com: Repick this patch for unwinding userstack backtrace by perf and libunwind on MIPS based CPU.] [ralf@linux-mips.org: Add perf_get_regs_user() which is required after 'commit 88a7c26a ("perf: Move task_pt_regs sampling into arch code")'.] [yangtiezhu@loongson.cn: Fix build error about perf_get_regs_user() after commit 76a4efa8 ("perf/arch: Remove perf_sample_data::regs_user_copy"), and also separate the original patches into two parts (MIPS kernel and perf tools) to merge easily.] The original patches: https://lore.kernel.org/patchwork/patch/1126521/ https://lore.kernel.org/patchwork/patch/1126520/Signed-off-by:
David Daney <david.daney@cavium.com> Signed-off-by:
Ralf Baechle <ralf@linux-mips.org> Signed-off-by:
Archer Yan <ayan@wavecomp.com> Signed-off-by:
Tiezhu Yang <yangtiezhu@loongson.cn> Signed-off-by:
-
Alexander Lobakin authored
Since commit 02bd530f ("MIPS: generic: Increase NR_IRQS to 256") include/asm/mach-pistachio/irq.h just does nothing. Remove the file along with mach-pistachio folder and include compiler directive. Signed-off-by:
-
Bert Vermeulen authored
Signed-off-by:
Bert Vermeulen <bert@biot.com> Signed-off-by:
-
Bert Vermeulen authored
Signed-off-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Bert Vermeulen authored
This is just enough system to boot the kernel with earlycon working. Signed-off-by:
Sander Vanheule <sander@svanheule.net> Signed-off-by:
-
Bert Vermeulen authored
The RTL838x/839x family of SoCs are Realtek switches with an embedded MIPS core. * RTL838x - 500MHz 4kce single core - 1Gbit ports and L2 features * RTL839x - 700MHz 34Kc single core - 1Gbit ports and L2 features These switches, depending on the exact part number, will have anywhere between 8 and 52 ports. The MIPS core is wired to a switch cpu port which has a tagging feature allowing us to make use of the DSA subsystem. The SoCs are somewhat basic in certain areas, getting better with more advanced features on newer series. The switch functionality is MMIO-mapped via a large MFD region. The SoCs have the following peripherals * ethernet * switch * uart - ns16550a * spi-flash interface * gpio * wdt * led The code was derived from various vendor SDKs based on Linux v2.6 kernels. This patchset allows us to boot RTL838x/RTL839x units with basic support. Most of the other drivers are already written and functional, and work to get them upstream is already in progress. Signed-off-by:
Birger Koblitz <mail@birger-koblitz.de> Signed-off-by:
John Crispin <john@phrozen.org> Signed-off-by:
-
Bert Vermeulen authored
Signed-off-by:
-
Sander Vanheule authored
Reset lines with indices smaller than 8 are currently considered invalid by the rt2880-reset reset controller. The MT7621 SoC uses a number of these low reset lines. The DTS defines reset lines "hsdma", "fe", and "mcm" with respective values 5, 6, and 2. As a result of the above restriction, these resets cannot be asserted or de-asserted by the reset controller. In cases where the bootloader does not de-assert these lines, this results in e.g. the MT7621's internal switch staying in reset. Change the reset controller to only ignore the system reset, so all reset lines with index greater than 0 are considered valid. Signed-off-by:
-
Thomas Bogendoerfer authored
Selection of the DTB to be used was burried in more or less readable code in head.S. Move this code into a inline helper function and use it. Signed-off-by:
Florian Fainelli <f.fainelli@gmail.com>
-
- 03 Feb, 2021 4 commits
-
-
Bhaskar Chowdhury authored
s/logcal/logical/ s/intercpu/inter-CPU/ Signed-off-by:
Bhaskar Chowdhury <unixbhaskar@gmail.com> Acked-by:
Randy Dunlap <rdunlap@infradead.org> Signed-off-by:
-
Yang Li authored
Eliminate the following coccicheck warning: ./arch/mips/mti-malta/malta-time.c:141:2-3: Unneeded semicolon Reported-by:
Abaci Robot <abaci@linux.alibaba.com> Signed-off-by:
Yang Li <yang.lee@linux.alibaba.com> Signed-off-by:
-
Yang Li authored
Eliminate the following coccicheck warning: ./arch/mips/kvm/mips.c:151:2-3: Unneeded semicolon Reported-by:
Huacai Chen <chenhuacai@kernel.org> Signed-off-by:
-
Huang Pei authored
MIPS always save kernel stack pointer in regs[29] Signed-off-by:
Huang Pei <huangpei@loongson.cn> Signed-off-by:
-
- 27 Jan, 2021 6 commits
-
-
Tiezhu Yang authored
Fix the following build error when make M=samples/bpf used with Clang: CLANG-bpf samples/bpf/sockex2_kern.o In file included from samples/bpf/sockex2_kern.c:7: In file included from ./include/uapi/linux/if_tunnel.h:7: In file included from ./include/linux/ip.h:16: In file included from ./include/linux/skbuff.h:28: In file included from ./include/net/checksum.h:22: ./arch/mips/include/asm/checksum.h:161:9: error: unsupported inline asm: input with type 'unsigned long' matching output with type '__wsum' (aka 'unsigned int') : "0" ((__force unsigned long)daddr), ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1 error generated. This is a known issue on MIPS [1], the changed code can be compiled successfully by both GCC and Clang. [1] https://lore.kernel.org/linux-mips/CAG_fn=W0JHf8QyUX==+rQMp8PoULHrsQCa9Htffws31ga8k-iw@mail.gmail.com/Signed-off-by: -
Yanteng Si authored
MIPS can now use the default uncached_access like other archs. Signed-off-by:
Yanteng Si <siyanteng@loongson.cn> Acked-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
-
Yanteng Si authored
We no longer need the MESA workaround, so remove it. Signed-off-by:
-
Chengyang Fan authored
Remove a superfluous semicolon after function definition. Signed-off-by:
Chengyang Fan <cy.fan@huawei.com> Signed-off-by:
-
Thomas Bogendoerfer authored
I couldn't find any user of the dubious vpe_getcwd so far. So remove it and get rid of another set_fs(KERNEL_DS). Signed-off-by: -
Souptick Joarder authored
kernel test robot throws below warning -> arch/mips/kernel/cacheinfo.c:112:3: warning: Variable 'level' is modified but its new value is never used. [unreadVariable] Remove unnecessary increment of level at the end. Reported-by:
Souptick Joarder <jrdr.linux@gmail.com> Reviewed-by:
Jiaxun Yang <jiaxun.yang@flygoat.com> Signed-off-by:
-
- 25 Jan, 2021 6 commits
-
-
Jinyang He authored
For those leaf functions, they are likely to have no stack operations. Add is_jr_ra_ins() to determine whether jr ra has been touched before the frame_size is found. Without this patch, the get frame_size operation may be out of range and get the frame_size from the next nested function. There is no POOL32A format in uapi/asm/inst.h, so some bits here use the format of r_format instead. e.g. --------------------------------------------------------------------- | format | 31:26 | 25:21 | 20:16 | 15:6 | 5:0 | -----------------+---------+-------+-------+------------+------------ | pool32a_format | pool32a | rt | rs | jalrc | pool32axf | -----------------+---------+-------+-------+------------+------------ | r_format | opcode | rs | rt | rd:5, re:5 | func | --------------------------------------------------------------------- Signed-off-by:
-
Jinyang He authored
[1]: Commit b6c7a324 ("MIPS: Fix get_frame_info() handling of microMIPS function size") [2]: Commit 2b424cfc ("MIPS: Remove function size check in get_frame_info()") First patch added a constant to check the number of iterations against. Second patch fixed the situation that info->func_size is zero. However, func_size member became useless after the second commit. Without ip_end, the get frame_size operation may be out of range although KALLSYMS enabled. Thus, check func_size first. Then make ip_end be the sum of ip and a constant (512) if func_size is equal to 0. Otherwise make ip_end be the sum of ip and func_size. Signed-off-by:
-
Jinyang He authored
mm16_r5_format.rt is 5 bits, so directly judge the value if equal or not. mm_jalr_op requires 7th to 16th bits. These 10 which bits generated by shifting u_format.uimmediate by 6 may be affected by sign extension. Thus, take out the 10 bits for comparison. Without this patch, errors may occur, such as these bits are all ones. Signed-off-by:
-
Jinyang He authored
Some headers are not necessary, remove them and sort includes. Signed-off-by:
-
Maciej W. Rozycki authored
I find linux-mips.org too unreliable to rely on, so move to a place I have proper control over. Signed-off-by:
Maciej W. Rozycki <macro@orcam.me.uk> Signed-off-by:
-
Lauri Kasanen authored
This adds support for the Nintendo 64 console's sound. Signed-off-by:
Lauri Kasanen <cand@gmx.com> Reviewed-by:
Takashi Iwai <tiwai@suse.de> Signed-off-by:
-
- 22 Jan, 2021 7 commits
-
-
Arnd Bergmann authored
kgdb fails to build when the FPU support is disabled: arch/mips/kernel/kgdb.c: In function 'dbg_set_reg': arch/mips/kernel/kgdb.c:147:35: error: 'struct thread_struct' has no member named 'fpu' 147 | memcpy((void *)¤t->thread.fpu.fcr31, mem, | ^ arch/mips/kernel/kgdb.c:155:34: error: 'struct thread_struct' has no member named 'fpu' 155 | memcpy((void *)¤t->thread.fpu.fpr[fp_reg], mem, This is only relevant for CONFIG_EXPERT=y, so disallowing it in Kconfig is an easier workaround than fixing it properly. Reported-by:Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Arnd Bergmann authored
The kernel test robot keeps reporting the same bug when it shows up in new files after random unrelated patches: In file included from arch/mips/include/uapi/asm/byteorder.h:13, from arch/mips/include/asm/bitops.h:20, from include/linux/bitops.h:26, from include/linux/kernel.h:12, from include/linux/clk.h:13, from drivers/base/regmap/regmap-mmio.c:7: include/linux/byteorder/big_endian.h:8:2: warning: #warning inconsistent configuration, needs CONFIG_CPU_BIG_ENDIAN [-Wcpp] 8 | #warning inconsistent configuration, needs CONFIG_CPU_BIG_ENDIAN | ^~~~~~~ drivers/base/regmap/regmap-mmio.c: In function 'regmap_mmio_gen_context': >> drivers/base/regmap/regmap-mmio.c:274:2: error: duplicate case value 274 | case REGMAP_ENDIAN_NATIVE: | ^~~~ drivers/base/regmap/regmap-mmio.c:246:2: note: previously used here 246 | case REGMAP_ENDIAN_NATIVE: The problem is that some randconfig builds end up on the MIPS jazz platform with neither CONFIG_CPU_BIG_ENDIAN nor CONFIG_CPU_LITTLE_ENDIAN because no specific machine is selected. As it turns out, all jazz machines support little-endian kernels, so this can simply be allowed globally. Reported-by: -
Tiezhu Yang authored
The kernel definitions of MIPSInst_FMA_FUNC and MIPSInst_FMA_FFMT are not consistent with MADD.fmt, NMADD.fmt and NMSUB.fmt in the MIPS64 manual [1], the field func is bit 5..3 and fmt is bit 2..0, fix them. Otherwise there exists error when add new instruction simulation. [1] https://www.mips.com/?do-download=the-mips64-instruction-set-v6-06Reported-by:
Ming Wang <wangming01@loongson.cn> Signed-off-by:
-
Lauri Kasanen authored
Add support for the Nintendo 64. Signed-off-by:
-
Lauri Kasanen authored
This reverts commit f9065b54. We're adding Nintendo 64 support, so the VR4300 is no longer unused. Signed-off-by:
-
Yanteng Si authored
This commit adds a prototype to fix error at W=1: arch/mips/mm/cache.c:129:6: error: no previous prototype for '__update_cache' [-Werror=missing-prototypes] Signed-off-by:
-
Thomas Bogendoerfer authored
When there's a fatal signal pending, MIPS's do_page_fault() implementation returns. The intent is that we'll return to the faulting userspace instruction, delivering the signal on the way. However, if we take a fatal signal during fixing up a uaccess, this results in a return to the faulting kernel instruction, which will be instantly retried, resulting in the same fault being taken forever. As the task never reaches userspace, the signal is not delivered, and the task is left unkillable. While the task is stuck in this state, it can inhibit the forward progress of the system. To avoid this, we must ensure that when a fatal signal is pending, we apply any necessary fixup for a faulting kernel instruction. Thus we will return to an error path, and it is up to that code to make forward progress towards delivering the fatal signal. [ Description taken from commit 746a272e ("ARM: 8692/1: mm: abort uaccess retries upon fatal signal") ] Signed-off-by:
Mark Rutland <mark.rutland@arm.com>
-
- 19 Jan, 2021 2 commits
-
-
Yanteng Si authored
This commit adds a prototype to fix error at W=1: arch/mips/kernel/irq.c:52:13: error: no previous prototype for 'init_IRQ' [-Werror=missing-prototypes] Signed-off-by:
-
Kevin Hao authored
This reverts commit d9df9fb9. For the OCTEON boards, it need to patch the built-in DTB before using it. Previously it judges if it is a built-in DTB by checking fw_passed_dtb. But after commit 37e5c69f ("MIPS: head.S: Init fw_passed_dtb to builtin DTB", the fw_passed_dtb is initialized even when using built-in DTB. This causes the OCTEON boards boot broken due to an unpatched built-in DTB is used. Revert the commit d9df9fb9 to restore the codes before the fw_passed_dtb is used and then fix this issue. Fixed: 37e5c69f ("MIPS: head.S: Init fw_passed_dtb to builtin DTB") Cc: stable@vger.kernel.org Suggested-by:
Kevin Hao <haokexin@gmail.com> Signed-off-by:
-
- 18 Jan, 2021 1 commit
-
-
Alexander Lobakin authored
For now, vmlinux relocation functions for relocatable kernel are implemented as an array of handlers of a particular type. Convert that array into a single switch-case function to: - remove unused arguments; - change the return type of simple handlers to void; - remove the array and don't use any data at all; - avoid using indirect calls; - allow the compiler to inline and greatly optimize the relocation function[s]; and also mark do_relocations() and show_kernel_relocation() static as they aren't used anywhere else. The result on MIPS32 R2 with GCC 10.2 -O2 is: scripts/bloat-o-meter -c arch/mips/kernel/__relocate.o arch/mips/kernel/relocate.o add/remove: 0/6 grow/shrink: 1/0 up/down: 356/-640 (-284) Function old new delta relocate_kernel 852 1208 +356 apply_r_mips_32_rel 20 - -20 apply_r_mips_hi16_rel 40 - -40 apply_r_mips_64_rel 44 - -44 apply_r_mips_26_rel 144 - -144 show_kernel_relocation 164 - -164 do_relocations 228 - -228 Total: Before=1780, After=1496, chg -15.96% add/remove: 0/1 grow/shrink: 0/0 up/down: 0/-76 (-76) Data old new delta reloc_handlers_rel 76 - -76 Total: Before=92, After=16, chg -82.61% add/remove: 0/0 grow/shrink: 0/0 up/down: 0/0 (0) RO Data old new delta Total: Before=0, After=0, chg +0.00% All functions were collapsed into the main one, relocate_kernel(). Signed-off-by:
-
