- 16 Aug, 2023 34 commits
-
-
Benjamin Gray authored
Now that ptrace and perf are no longer exclusive, update the test to exercise interesting interactions. An assembly file is used for the children to allow precise instruction choice and addresses, while avoiding any compiler quirks. Signed-off-by:
Benjamin Gray <bgray@linux.ibm.com> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Link: https://msgid.link/20230801011744.153973-7-bgray@linux.ibm.com
-
Benjamin Gray authored
ptrace and perf watchpoints were considered incompatible in commit 29da4f91 ("powerpc/watchpoint: Don't allow concurrent perf and ptrace events"), but the logic in that commit doesn't really apply. Ptrace doesn't automatically single step; the ptracer must request this explicitly. And the ptracer can do so regardless of whether a ptrace/perf watchpoint triggered or not: it could single step every instruction if it wanted to. Whatever stopped the ptracee before executing the instruction that would trigger the perf watchpoint is no longer relevant by this point. To get correct behaviour when perf and ptrace are watching the same data we must ignore the perf watchpoint. After all, ptrace has before-execute semantics, and perf is after-execute, so perf doesn't actually care about the watchpoint trigger at this point in time. Pausing before execution does not mean we will actually end up executing the instruction. Importantly though, we don't remove the perf watchpoint yet. This is key. The ptracer is free to do whatever it likes right now. E.g., it can continue the process, single step. or even set the child PC somewhere completely different. If it does try to execute the instruction though, without reinserting the watchpoint (in which case we go back to the start of this example), the perf watchpoint would immediately trigger. This time there is no ptrace watchpoint, so we can safely perform a single step and increment the perf counter. Upon receiving the single step exception, the existing code already handles propagating or consuming it based on whether another subsystem (e.g. ptrace) requested a single step. Again, this is needed with or without perf/ptrace exclusion, because ptrace could be single stepping this instruction regardless of if a watchpoint is involved. Signed-off-by:
-
Benjamin Gray authored
We only remove watchpoints when they have the perf_single_step flag set, so we can reinsert them during the first iteration. Signed-off-by:
-
Benjamin Gray authored
There is a bug in the current watchpoint tracking logic, where the teardown in arch_unregister_hw_breakpoint() uses bp->ctx->task, which it does not have a reference of and parallel threads may be in the process of destroying. This was partially addressed in commit fb822e60 ("powerpc/hw_breakpoint: Fix oops when destroying hw_breakpoint event"), but the underlying issue of accessing a struct member in an unknown state still remained. Syzkaller managed to trigger a null pointer derefernce due to the race between the task destructor and checking the pointer and dereferencing it in the loop. While this null pointer dereference could be fixed by using READ_ONCE to access the task up front, that just changes the error to manipulating possbily freed memory. Instead, the breakpoint logic needs to be reworked to remove any dependency on a context or task struct during breakpoint removal. The reason we have this currently is to clear thread.last_hit_ubp. This member is used to differentiate the perf DAWR single-step sequence from other causes of single-step, such as userspace just calling ptrace(PTRACE_SINGLESTEP, ...). We need to differentiate them because, when the single step interrupt is received, we need to know whether to re-insert the DAWR breakpoint (perf) or not (ptrace / other). arch_unregister_hw_breakpoint() needs to clear this information to prevent dangling pointers to possibly freed memory. These pointers are dereferenced in single_step_dabr_instruction() without a way to check their validity. This patch moves the tracking of this information to the breakpoint itself. This means we no longer have to do anything special to clean up. Signed-off-by:
-
Benjamin Gray authored
info is cheap to retrieve, and is likely optimised by the compiler anyway. On the other hand, propagating it across the functions makes it possible to be inconsistent and adds needless complexity. Remove it, and invoke counter_arch_bp() when we need to work with it. As we don't persist it, we just use the local bp array to track whether we are ignoring a breakpoint. Signed-off-by:
-
Benjamin Gray authored
The behaviour of the thread_change_pc() function is a bit cryptic without being more familiar with how the watchpoint logic handles perf's after-execute semantics. Expand the comment to explain why we can re-insert the breakpoint and unset the perf_single_step flag. Signed-off-by:
-
Kajol Jain authored
docs: ABI: sysfs-bus-event_source-devices-hv_gpci: Document affinity_domain_via_partition sysfs interface file Add details of the new hv-gpci interface file called "affinity_domain_via_partition" in the ABI documentation. Reviewed-by:
Athira Rajeev <atrajeev@linux.vnet.ibm.com> Signed-off-by:
Kajol Jain <kjain@linux.ibm.com> Signed-off-by:
-
Kajol Jain authored
powerpc/hv_gpci: Add sysfs file inside hv_gpci device to show affinity domain via partition information The hcall H_GET_PERF_COUNTER_INFO with counter request value as AFFINITY_DOMAIN_INFORMATION_BY_PARTITION(0XB1), can be used to get the system affinity domain via partition information. To expose the system affinity domain via partition information, patch adds sysfs file called "affinity_domain_via_partition" to the "/sys/devices/hv_gpci/interface/" of hv_gpci pmu driver. Add new entry for AFFINITY_DOMAIN_VIA_PAR in sysinfo_counter_request array, which points to the counter request value "affinity_domain_via_partition" in hv-gpci.c file. Also add a new function called "affinity_domain_via_partition_result_parse" to parse the hcall result and store it in output buffer. The affinity_domain_via_partition sysfs file is only available for power10 and above platforms. Add a macro called INTERFACE_AFFINITY_DOMAIN_VIA_PAR_ATTR, which points to the index of NULL placeholder, for affinity_domain_via_partition attribute in interface_attrs array. Also updated the value of INTERFACE_NULL_ATTR macro in hv-gpci.c file. Reviewed-by:
-
Kajol Jain authored
docs: ABI: sysfs-bus-event_source-devices-hv_gpci: Document affinity_domain_via_domain sysfs interface file Add details of the new hv-gpci interface file called "affinity_domain_via_domain" in the ABI documentation. Reviewed-by:
-
Kajol Jain authored
powerpc/hv_gpci: Add sysfs file inside hv_gpci device to show affinity domain via domain information The hcall H_GET_PERF_COUNTER_INFO with counter request value as AFFINITY_DOMAIN_INFORMATION_BY_DOMAIN(0XB0), can be used to get the system affinity domain via domain information. To expose the system affinity domain via domain information, patch adds sysfs file called "affinity_domain_via_domain" to the "/sys/devices/hv_gpci/interface/" of hv_gpci pmu driver. Add new entry for AFFINITY_DOMAIN_VIA_DOM in sysinfo_counter_request array, which points to the counter request value "affinity_domain_via_domain" in hv-gpci.c file. The affinity_domain_via_domain sysfs file is only available for power10 and above platforms. Add a macro called INTERFACE_AFFINITY_DOMAIN_VIA_DOM_ATTR, which points to the index of NULL placeholder, for affinity_domain_via_domain attribute in interface_attrs array. Also updated the value of INTERFACE_NULL_ATTR macro in hv-gpci.c file. Reviewed-by:
-
Kajol Jain authored
docs: ABI: sysfs-bus-event_source-devices-hv_gpci: Document affinity_domain_via_virtual_processor sysfs interface file Add details of the new hv-gpci interface file called "affinity_domain_via_virtual_processor" in the ABI documentation. Reviewed-by:
-
Kajol Jain authored
powerpc/hv_gpci: Add sysfs file inside hv_gpci device to show affinity domain via virtual processor information The hcall H_GET_PERF_COUNTER_INFO with counter request value as AFFINITY_DOMAIN_INFORMATION_BY_VIRTUAL_PROCESSOR(0XA0), can be used to get the system affinity domain via virtual processor information. To expose the system affinity domain via virtual processor information, patch adds sysfs file called "affinity_domain_via_virtual_processor" to the "/sys/devices/hv_gpci/interface/" of hv_gpci pmu driver. The affinity_domain_via_virtual_processor sysfs file is only available for power10 and above platforms. Add a macro called INTERFACE_AFFINITY_DOMAIN_VIA_VP_ATTR, which points to the index of NULL placeholder, for affinity_domain_via_virtual_processor attribute in interface_attrs array. Also updated the value of INTERFACE_NULL_ATTR macro in hv-gpci.c file. Reviewed-by:
-
Kajol Jain authored
Add details of the new hv-gpci interface file called "processor_config" in the ABI documentation. Reviewed-by:
-
Kajol Jain authored
The hcall H_GET_PERF_COUNTER_INFO with counter request value as PROCESSOR_CONFIG(0X90), can be used to get the system processor configuration information. To expose the system processor config information, patch adds sysfs file called "processor_config" to the "/sys/devices/hv_gpci/interface/" of hv_gpci pmu driver. Add enum and sysinfo_counter_request array to get required counter request value in hv-gpci.c file. Also add a new function called "sysinfo_device_attr_create", which will create and return required device attribute to the add_sysinfo_interface_files function. The processor_config sysfs file is only available for power10 and above platforms. Add a new macro called INTERFACE_PROCESSOR_CONFIG_ATTR, which points to the index of NULL placefolder, for processor_config attribute in the interface_attrs array. Also add macro INTERFACE_NULL_ATTR which points to index of NULL attribute in interface_attrs array. Reviewed-by:
-
Kajol Jain authored
docs: ABI: sysfs-bus-event_source-devices-hv_gpci: Document processor_bus_topology sysfs interface file Add details of the new hv-gpci interface file called "processor_bus_topology" in the ABI documentation. Reviewed-by:
-
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:
-
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 6 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:
-
