- 16 Aug, 2023 19 commits
-
-
Kajol Jain authored
The hcall H_GET_PERF_COUNTER_INFO with counter request value as PROCESSOR_BUS_TOPOLOGY(0XD0), can be used to get the system topology information. To expose the system topology information, patch adds sysfs file called "processor_bus_topology" to the "/sys/devices/hv_gpci/interface/" of hv_gpci pmu driver. Add macro for PROCESSOR_BUS_TOPOLOGY counter request value in hv-gpci.c file. Also add a new function called "systeminfo_gpci_request", to make the H_GET_PERF_COUNTER_INFO hcall with added macro and populates the output buffer. The processor_bus_topology sysfs file is only available for power10 and above platforms. Add a new function called "add_sysinfo_interface_files", which will add processor_bus_topology attribute in the interface_attrs array, only for power10 and above platforms. Also add macro INTERFACE_PROCESSOR_BUS_TOPOLOGY_ATTR in hv-gpci.c file, which points to the index of NULL placefolder, for processor_bus_topology attribute. Reviewed-by:
Athira Rajeev <atrajeev@linux.vnet.ibm.com> Signed-off-by:
Kajol Jain <kjain@linux.ibm.com> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Link: https://msgid.link/20230729073455.7918-2-kjain@linux.ibm.com
-
Linus Walleij authored
Making virt_to_pfn() a static inline taking a strongly typed (const void *) makes the contract of a passing a pointer of that type to the function explicit and exposes any misuse of the macro virt_to_pfn() acting polymorphic and accepting many types such as (void *), (unitptr_t) or (unsigned long) as arguments without warnings. Move the virt_to_pfn() and related functions below the declaration of __pa() so it compiles. For symmetry do the same with pfn_to_kaddr(). As the file is included right into the linker file, we need to surround the functions with ifndef __ASSEMBLY__ so we don't cause compilation errors. The conversion moreover exposes the fact that pmd_page_vaddr() was returning an unsigned long rather than a const void * as could be expected, so all the sites defining pmd_page_vaddr() had to be augmented as well. Finally the KVM code in book3s_64_mmu_hv.c was passing an unsigned int to virt_to_phys() so fix that up with a cast so the result compiles. Signed-off-by:
Linus Walleij <linus.walleij@linaro.org> [mpe: Fixup kfence.h, simplify pfn_to_kaddr() & pmd_page_vaddr()] Signed-off-by:
-
Xiongfeng Wang authored
PCI core API pci_dev_id() can be used to get the BDF number for a pci device. We don't need to compose it mannually. Use pci_dev_id() to simplify the code a little bit. Signed-off-by:
Xiongfeng Wang <wangxiongfeng2@huawei.com> Reviewed-by:
Frederic Barrat <fbarrat@linux.ibm.com> Signed-off-by:
-
Gautam Menghani authored
Remove an unnecessary piece of code that does an endianness conversion but does not use the result. The following warning was reported by Clang's static analyzer: arch/powerpc/sysdev/xics/ics-opal.c:114:2: warning: Value stored to 'server' is never read [deadcode.DeadStores] server = be16_to_cpu(oserver); 'server' was used as a parameter to opal_get_xive() in commit 5c7c1e94 ("powerpc/powernv: Add OPAL ICS backend") when it was introduced. 'server' was also used in an error message for the call to opal_get_xive(). 'server' was always later set by a call to ics_opal_mangle_server() before being used. Commit bf8e0f89 ("powerpc/powernv: Fix endian issues in OPAL ICS backend") used a new variable 'oserver' as the parameter to opal_get_xive() instead of 'server' for endian correctness. It also removed 'server' from the error message for the call to opal_get_xive(). Fix the warning by removing the server variable assignment. Fixes: bf8e0f89 ("powerpc/powernv: Fix endian issues in OPAL ICS backend") Reviewed-by:
Jordan Niethe <jniethe5@gmail.com> Signed-off-by:
Gautam Menghani <gautam@linux.ibm.com> Signed-off-by:
-
ruanjinjie authored
If device_register() returns error in ibmebus_bus_init(), name of kobject which is allocated in dev_set_name() called in device_add() is leaked. As comment of device_add() says, it should call put_device() to drop the reference count that was set in device_initialize() when it fails, so the name can be freed in kobject_cleanup(). Signed-off-by:
ruanjinjie <ruanjinjie@huawei.com> Signed-off-by:
-
Masahiro Yamada authored
All *.S files under arch/powerpc/ have been converted to include <linux/export.h> instead of <asm/export.h>. Remove <asm/export.h>. Signed-off-by:
Masahiro Yamada <masahiroy@kernel.org> Signed-off-by:
-
Masahiro Yamada authored
Commit ddb5cdba ("kbuild: generate KSYMTAB entries by modpost") deprecated <asm/export.h>, which is now a wrapper of <linux/export.h>. Replace #include <asm/export.h> with #include <linux/export.h>. After all the <asm/export.h> lines are converted, <asm/export.h> and <asm-generic/export.h> will be removed. Signed-off-by:
-
Masahiro Yamada authored
There is no EXPORT_SYMBOL line there, hence #include <asm/export.h> is unneeded. Signed-off-by:
-
Nick Desaulniers authored
Clang didn't recognize the instruction tlbilxlpid. This was fixed in clang-18 [0] then backported to clang-17 [1]. To support clang-16 and older, rather than using that instruction bare in inline asm, add it to ppc-opcode.h and use that macro as is done elsewhere for other instructions. Link: https://github.com/ClangBuiltLinux/linux/issues/1891 Link: https://github.com/llvm/llvm-project/issues/64080 Link: https://github.com/llvm/llvm-project/commit/53648ac1d0c953ae6d008864dd2eddb437a92468 [0] Link: https://github.com/llvm/llvm-project-release-prs/commit/0af7e5e54a8c7ac665773ac1ada328713e8338f5 [1] Reported-by:
kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/llvm/202307211945.TSPcyOhh-lkp@intel.com/Suggested-by:
Nick Desaulniers <ndesaulniers@google.com> Reviewed-by:
Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by:
-
Xiongfeng Wang authored
PCI core add pci_find_vsec_capability() to query VSEC. We can use that core API to simplify the code. The only logical change is that pci_find_vsec_capability check the Vendor ID before finding the VSEC. PCI spec rev 5.0 says in 7.9.5.2 Vendor-Specific Header: VSEC ID - This field is a vendor-defined ID number that indicates the nature and format of the VSEC structure Software must qualify the Vendor ID before interpreting this field. Signed-off-by:
Andrew Donnellan <ajd@linux.ibm.com> Signed-off-by:
-
Christophe Leroy authored
That ifdef was introduced by commit 1458dd95 ("powerpc/8xx: Handle CPU6 ERRATA directly in mtspr() macro") and left over by commit 2a45addd ("powerpc/8xx: Remove CPU6 ERRATA Workaround") Remove it. Signed-off-by:
-
Christophe Leroy authored
objtool reports two folliwng warnings: arch/powerpc/lib/sstep.o: warning: objtool: copy_mem_out+0x3c (.text+0x30c): call to __copy_mem_out() with UACCESS enabled arch/powerpc/lib/sstep.o: warning: objtool: emulate_dcbz+0x70 (.text+0x4dc): call to __emulate_dcbz() with UACCESS enabled Mark __copy_mem_out() and __emulate_dcbz() __always_inline Signed-off-by: -
Christophe Leroy authored
Since commit 449012da ("[POWERPC] cpm2: Infrastructure code cleanup.") cpm2_map() is just returning cpm2_immr pointer and cpm2_unmap() does nothing. We already have parts of code that use cpm2_immr directly so get rid of cpm2_map() and cpm2_unmap() by using cpm2_immr directly. And avoid going through local pointers that hide the pointed structure. Signed-off-by:
-
Christophe Leroy authored
Since commit fb533d0c ("[POWERPC] 8xx: Infrastructure code cleanup.") immr_map() is just returning mpc8xxx_immr pointer and immr_unmap() do nothing. We already have parts of code that use mpc8xxx_immr directly so get rid of immr_map() and immr_unmap() by using mpc8xxx_immr directly. And avoid going through local pointers that hide the pointed structure. Signed-off-by:
-
Christophe Leroy authored
CONFIG_PCI_8260 is not used anymore, remove it. Signed-off-by:
-
Christophe Leroy authored
SIU_INT_IRQ1 is not used anywhere and __IO_BASE is defined in asm/io.h Remove m82xx_pci.h Then the only thing remaining in mpc8260.h is MPC82XX_BCR_PLDP Move MPC82XX_BCR_PLDP into asm/cpm2.h then remove mpc8260.h Signed-off-by:
-
Christophe Leroy authored
Do the same as for cmp2_immr : declare it at the same place as its type immap_t, that is in 8xx_immap.h instead of fs_pd.h Signed-off-by:
-
Christophe Leroy authored
tqm8xx_setup.c and fs_enet.h don't use any items provided by fs_pd.h Remove unneeded #include <asm/fs_pd.h> Signed-off-by:
-
Zheng Zengkai authored
PCI core API pci_dev_id() can be used to get the BDF number for a pci device. We don't need to compose it mannually. Use pci_dev_id() to simplify the code a little bit. Signed-off-by:
Zheng Zengkai <zhengzengkai@huawei.com> Acked-by:
-
- 14 Aug, 2023 7 commits
-
-
Arnd Bergmann authored
This is the module init function, which by definition is used only locally, so mark it static to avoid a warning: drivers/macintosh/ams/ams-core.c:179:12: error: no previous prototype for 'ams_init' [-Werror=missing-prototypes] Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Reviewed-by:
-
Randy Dunlap authored
Don't use kernel-doc "/**" comment format for non-kernel-doc comments. This prevents a kernel-doc warning: arch/powerpc/platforms/pseries/plpks.c:186: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst * Label is combination of label attributes + name. Fixes: 2454a7af ("powerpc/pseries: define driver for Platform KeyStore") Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Reported-by:
-
Christophe Leroy authored
With skiboot_defconfig, Clang reports: CC arch/powerpc/mm/book3s64/radix_tlb.o arch/powerpc/mm/book3s64/radix_tlb.c:419:20: error: unused function '_tlbie_pid_lpid' [-Werror,-Wunused-function] static inline void _tlbie_pid_lpid(unsigned long pid, unsigned long lpid, ^ arch/powerpc/mm/book3s64/radix_tlb.c:663:20: error: unused function '_tlbie_va_range_lpid' [-Werror,-Wunused-function] static inline void _tlbie_va_range_lpid(unsigned long start, unsigned long end, ^ This is because those functions are only called from functions enclosed in a #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE Move below functions inside that #ifdef * __tlbie_pid_lpid(unsigned long pid, * __tlbie_va_lpid(unsigned long va, unsigned long pid, * fixup_tlbie_pid_lpid(unsigned long pid, unsigned long lpid) * _tlbie_pid_lpid(unsigned long pid, unsigned long lpid, * fixup_tlbie_va_range_lpid(unsigned long va, * __tlbie_va_range_lpid(unsigned long start, unsigned long end, * _tlbie_va_range_lpid(unsigned long start, unsigned long end, Fixes: f0c6fbbb ("KVM: PPC: Book3S HV: Add support for H_RPT_INVALIDATE") Reported-by: -
Arnd Bergmann authored
Randconfig testing with W=1 showed up these warnings that I'd like to enable by default: arch/powerpc/xmon/xmon.c: In function 'dump_tlb_book3e': arch/powerpc/xmon/xmon.c:3833:42: error: variable 'lrat' set but not used [-Werror=unused-but-set-variable] 3833 | int i, tlb, ntlbs, pidsz, lpidsz, rasz, lrat = 0; | ^~~~ arch/powerpc/xmon/xmon.c:3831:23: error: variable 'lpidmask' set but not used [-Werror=unused-but-set-variable] 3831 | u32 mmucfg, pidmask, lpidmask; | ^~~~~~~~ arch/powerpc/xmon/xmon.c:3831:14: error: variable 'pidmask' set but not used [-Werror=unused-but-set-variable] 3831 | u32 mmucfg, pidmask, lpidmask; | ^~~~~~~ Just remove these as they have been unused since the code was added in 2010. Fixes: 03247157 ("powerpc/book3e: Add TLB dump in xmon for Book3E") Signed-off-by: -
Arnd Bergmann authored
A while ago I created a2305e3d ("powerpc: mark local variables around longjmp as volatile") in order to allow building powerpc with -Wextra enabled on gcc-11. I tried this again with gcc-13 and found two more of the same issues, presumably based on slightly different optimization paths being taken here: arch/powerpc/xmon/xmon.c:3306:27: error: variable 'mm' might be clobbered by 'longjmp' or 'vfork' [-Werror=clobbered] arch/powerpc/kexec/crash.c:353:22: error: variable 'i' might be clobbered by 'longjmp' or 'vfork' [-Werror=clobbered] I checked a bunch of randconfigs and found only these two, so just address them the same way as the others. Signed-off-by:
-
Yuan Tan authored
Serial is a critical feature for logging and debuging, and the other architectures enable serial by default. Let's enable CONFIG_SERIAL_PMACZILOG and CONFIG_SERIAL_PMACZILOG_CONSOLE by default. Signed-off-by:
Yuan Tan <tanyuan@tinylab.org> Reviewed-by:
-
Michael Ellerman authored
Merge SMT changes we are sharing with the tip tree.
-
- 02 Aug, 2023 14 commits
-
-
Michael Ellerman authored
Integrate with the generic SMT support, so that when a CPU is DLPAR onlined it is brought up with the correct SMT mode. Signed-off-by:
-
Michael Ellerman authored
Add support for HOTPLUG_SMT, which enables the generic sysfs SMT support files in /sys/devices/system/cpu/smt, as well as the "nosmt" boot parameter. Implement the recently added hooks to allow partial SMT states, allow any number of threads per core. Tie the config symbol to HOTPLUG_CPU, which enables it on the major platforms that support SMT. If there are other platforms that want the SMT support that can be tweaked in future. Signed-off-by:
Laurent Dufour <ldufour@linux.ibm.com> Link: https://msgid.link/20230705145143.40545-10-ldufour@linux.ibm.com
-
Michael Ellerman authored
As part of the generic HOTPLUG_SMT code, there is support for disabling secondary SMT threads at boot time, by passing "nosmt" on the kernel command line. The way that is implemented is the secondary threads are brought partly online, and then taken back offline again. That is done to support x86 CPUs needing certain initialisation done on all threads. However powerpc has similar needs, see commit d70a54e2 ("powerpc/powernv: Ignore smt-enabled on Power8 and later"). For that to work the powerpc CPU hotplug callbacks need to be registered before secondary CPUs are brought online, otherwise __cpu_disable() fails due to smp_ops->cpu_disable being NULL. So split the basic initialisation into pseries_cpu_hotplug_init() which can be called early from setup_arch(). The DLPAR related initialisation can still be done later, because it needs to do allocations. Signed-off-by:
-
Michael Ellerman authored
Merge tag 'smp-core-for-ppc-23-07-28' of https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into topic/cpu-smt SMP core changes for powerpc
-
Laurent Dufour authored
Function name in the descriptor was not correct. Reported-by:
-
Uwe Kleine-König authored
Instead of resorting to BUG() ensure that the driver isn't unbound by suppressing its bind and unbind sysfs attributes. As the driver is built-in there is no way to remove a device once bound. As a nice side effect this allows to drop the remove function. Signed-off-by:
Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by:
-
Arnd Bergmann authored
There are a few warnings in powerpc64 defconfig builds after -Wmissing-prototypes gets promoted from W=1 to the default warning set: arch/powerpc/mm/book3s64/pgtable.c:422:6: error: no previous prototype for 'arch_report_meminfo' [-Werror=missing-prototypes] arch/powerpc/platforms/cell/ras.c:275:5: error: no previous prototype for 'cbe_sysreset_hack' [-Werror=missing-prototypes] arch/powerpc/platforms/cell/spu_manage.c:29:21: error: no previous prototype for 'spu_devnode' [-Werror=missing-prototypes] arch/powerpc/platforms/pasemi/time.c:12:17: error: no previous prototype for 'pas_get_boot_time' [-Werror=missing-prototypes] arch/powerpc/platforms/powermac/feature.c:1532:13: error: no previous prototype for 'g5_phy_disable_cpu1' [-Werror=missing-prototypes] arch/powerpc/platforms/86xx/pic.c:28:13: error: no previous prototype for 'mpc86xx_init_irq' [-Werror=missing-prototypes] drivers/pci/pci-sysfs.c:936:13: error: no previous prototype for 'pci_adjust_legacy_attr' [-Werror=missing-prototypes] Address these by including the right header files or marking the functions static. The audit.c one is a bit tricky since compat_audit.h cannot include regular kernel headers tht have conflicting types on 32-bit powerpc. Signed-off-by:
-
Benjamin Gray authored
While the target is volatile, the temporary variables used to access the target cast away the volatile. This is undefined behaviour, and a compiler may optimise away/reorder these accesses, breaking the test. This was observed with GCC 13.1.1, but it can be difficult to reproduce because of the dependency on compiler behaviour. Signed-off-by:
Benjamin Gray <bgray@linux.ibm.com> Signed-off-by:
-
Benjamin Gray authored
pid_max_addr() searches for the 'pid_max' symbol in /proc/kallsyms, and prints an error if it cannot find it. The error message has a typo, calling it pix_max. Signed-off-by:
-
Benjamin Gray authored
Many tests require specific hardware features/configurations that a typical machine might not have. As a result, it's common to see a test is skipped. But it is tedious to find out why a test is skipped when all it gives is the file location of the skip macro. Convert SKIP_IF() to SKIP_IF_MSG(), with appropriate descriptions of why the test is being skipped. This gives a general idea of why a test is skipped, which can be looked into further if it doesn't make sense. Signed-off-by:
-
Rob Herring authored
The DT of_device.h and of_platform.h date back to the separate of_platform_bus_type before it as merged into the regular platform bus. As part of that merge prepping Arm DT support 13 years ago, they "temporarily" include each other. They also include platform_device.h and of.h. As a result, there's a pretty much random mix of those include files used throughout the tree. In order to detangle these headers and replace the implicit includes with struct declarations, users need to explicitly include the correct includes. Signed-off-by:
Rob Herring <robh@kernel.org> [mpe: Fixup maple/setup.c which needs platform_device] Signed-off-by:
-
Nicholas Piggin authored
This performs lazy tlb mm shootdown when doing the exit TLB flush when all mm users go away and user mappings are removed, which avoids having to do the lazy tlb mm shootdown IPIs on the final mmput when all kernel references disappear. powerpc/64s uses a broadcast TLBIE for the exit TLB flush if remote CPUs need to be invalidated (unless TLBIE is disabled), so this doesn't necessarily save IPIs but it does avoid a broadcast TLBIE which is quite expensive. Signed-off-by:
Nicholas Piggin <npiggin@gmail.com> [mpe: Squash in preempt_disable/enable() fix from Nick] Signed-off-by:
-
Nicholas Piggin authored
When context switching away from an mm, add a CONFIG_DEBUG_VM warning check to ensure this CPU is still set in the mask. This could catch bugs where the mask is improperly trimmed while the CPU is still using the mm. Signed-off-by:
-
Nicholas Piggin authored
Avoid open-coded atomic_dec on mm->context.active_cpus and use the function made for it. Add CONFIG_DEBUG_VM underflow checking on the counter. Signed-off-by:
-
