Merge tag 'perf-core-2020-12-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull perf updates from Thomas Gleixner: "Core: - Better handling of page table leaves on archictectures which have architectures have non-pagetable aligned huge/large pages. For such architectures a leaf can actually be part of a larger entry. - Prevent a deadlock vs exec_update_mutex Architectures: - The related updates for page size calculation of leaf entries - The usual churn to support new CPUs - Small fixes and improvements all over the place" * tag 'perf-core-2020-12-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (24 commits) perf/x86/intel: Add Tremont Topdown support uprobes/x86: Fix fall-through warnings for Clang perf/x86: Fix fall-through warnings for Clang kprobes/x86: Fix fall-through warnings for Clang perf/x86/intel/lbr: Fix the return type of get_lbr_cycles() perf/x86/intel: Fix rtm_abort_event encoding on Ice Lake x86/kprobes: Restore BTF if the single-stepping is cancelled perf: Break deadlock involving exec_update_mutex sparc64/mm: Implement pXX_leaf_size() support powerpc/8xx: Implement pXX_leaf_size() support arm64/mm: Implement pXX_leaf_size() support perf/core: Fix arch_perf_get_page_size() mm: Introduce pXX_leaf_size() mm/gup: Provide gup_get_pte() more generic perf/x86/intel: Add event constraint for CYCLE_ACTIVITY.STALLS_MEM_ANY perf/x86/intel/uncore: Add Rocket Lake support perf/x86/msr: Add Rocket Lake CPU support perf/x86/cstate: Add Rocket Lake CPU support perf/x86/intel: Add Rocket Lake CPU support perf,mm: Handle non-page-table-aligned hugetlbfs ...
2020-12-14 17:34:12 -08:00
parent e857b6fcc5 c2208046bb
commit 8a8ca83ec3
21 changed files with 349 additions and 108 deletions
@@ -51,6 +51,8 @@
 #include <linux/proc_ns.h>
 #include <linux/mount.h>
 #include <linux/min_heap.h>
+#include <linux/highmem.h>
+#include <linux/pgtable.h>

 #include "internal.h"

@@ -1895,6 +1897,12 @@ static void __perf_event_header_size(struct perf_event *event, u64 sample_type)
 	if (sample_type & PERF_SAMPLE_CGROUP)
 		size += sizeof(data->cgroup);

+	if (sample_type & PERF_SAMPLE_DATA_PAGE_SIZE)
+		size += sizeof(data->data_page_size);
+
+	if (sample_type & PERF_SAMPLE_CODE_PAGE_SIZE)
+		size += sizeof(data->code_page_size);
+
 	event->header_size = size;
 }

@@ -6931,6 +6939,12 @@ void perf_output_sample(struct perf_output_handle *handle,
 	if (sample_type & PERF_SAMPLE_CGROUP)
 		perf_output_put(handle, data->cgroup);

+	if (sample_type & PERF_SAMPLE_DATA_PAGE_SIZE)
+		perf_output_put(handle, data->data_page_size);
+
+	if (sample_type & PERF_SAMPLE_CODE_PAGE_SIZE)
+		perf_output_put(handle, data->code_page_size);
+
 	if (sample_type & PERF_SAMPLE_AUX) {
 		perf_output_put(handle, data->aux_size);

@@ -6988,6 +7002,93 @@ static u64 perf_virt_to_phys(u64 virt)
 	return phys_addr;
 }

+/*
+ * Return the pagetable size of a given virtual address.
+ */
+static u64 perf_get_pgtable_size(struct mm_struct *mm, unsigned long addr)
+{
+	u64 size = 0;
+
+#ifdef CONFIG_HAVE_FAST_GUP
+	pgd_t *pgdp, pgd;
+	p4d_t *p4dp, p4d;
+	pud_t *pudp, pud;
+	pmd_t *pmdp, pmd;
+	pte_t *ptep, pte;
+
+	pgdp = pgd_offset(mm, addr);
+	pgd = READ_ONCE(*pgdp);
+	if (pgd_none(pgd))
+		return 0;
+
+	if (pgd_leaf(pgd))
+		return pgd_leaf_size(pgd);
+
+	p4dp = p4d_offset_lockless(pgdp, pgd, addr);
+	p4d = READ_ONCE(*p4dp);
+	if (!p4d_present(p4d))
+		return 0;
+
+	if (p4d_leaf(p4d))
+		return p4d_leaf_size(p4d);
+
+	pudp = pud_offset_lockless(p4dp, p4d, addr);
+	pud = READ_ONCE(*pudp);
+	if (!pud_present(pud))
+		return 0;
+
+	if (pud_leaf(pud))
+		return pud_leaf_size(pud);
+
+	pmdp = pmd_offset_lockless(pudp, pud, addr);
+	pmd = READ_ONCE(*pmdp);
+	if (!pmd_present(pmd))
+		return 0;
+
+	if (pmd_leaf(pmd))
+		return pmd_leaf_size(pmd);
+
+	ptep = pte_offset_map(&pmd, addr);
+	pte = ptep_get_lockless(ptep);
+	if (pte_present(pte))
+		size = pte_leaf_size(pte);
+	pte_unmap(ptep);
+#endif /* CONFIG_HAVE_FAST_GUP */
+
+	return size;
+}
+
+static u64 perf_get_page_size(unsigned long addr)
+{
+	struct mm_struct *mm;
+	unsigned long flags;
+	u64 size;
+
+	if (!addr)
+		return 0;
+
+	/*
+	 * Software page-table walkers must disable IRQs,
+	 * which prevents any tear down of the page tables.
+	 */
+	local_irq_save(flags);
+
+	mm = current->mm;
+	if (!mm) {
+		/*
+		 * For kernel threads and the like, use init_mm so that
+		 * we can find kernel memory.
+		 */
+		mm = &init_mm;
+	}
+
+	size = perf_get_pgtable_size(mm, addr);
+
+	local_irq_restore(flags);
+
+	return size;
+}
+
 static struct perf_callchain_entry __empty_callchain = { .nr = 0, };

 struct perf_callchain_entry *
@@ -7023,7 +7124,7 @@ void perf_prepare_sample(struct perf_event_header *header,

 	__perf_event_header__init_id(header, data, event);

-	if (sample_type & PERF_SAMPLE_IP)
+	if (sample_type & (PERF_SAMPLE_IP | PERF_SAMPLE_CODE_PAGE_SIZE))
 		data->ip = perf_instruction_pointer(regs);

 	if (sample_type & PERF_SAMPLE_CALLCHAIN) {
@@ -7142,6 +7243,17 @@ void perf_prepare_sample(struct perf_event_header *header,
 	}
 #endif

+	/*
+	 * PERF_DATA_PAGE_SIZE requires PERF_SAMPLE_ADDR. If the user doesn't
+	 * require PERF_SAMPLE_ADDR, kernel implicitly retrieve the data->addr,
+	 * but the value will not dump to the userspace.
+	 */
+	if (sample_type & PERF_SAMPLE_DATA_PAGE_SIZE)
+		data->data_page_size = perf_get_page_size(data->addr);
+
+	if (sample_type & PERF_SAMPLE_CODE_PAGE_SIZE)
+		data->code_page_size = perf_get_page_size(data->ip);
+
 	if (sample_type & PERF_SAMPLE_AUX) {
 		u64 size;

@@ -11720,24 +11832,6 @@ SYSCALL_DEFINE5(perf_event_open,
 		goto err_task;
 	}

-	if (task) {
-		err = mutex_lock_interruptible(&task->signal->exec_update_mutex);
-		if (err)
-			goto err_task;
-
-		/*
-		 * Preserve ptrace permission check for backwards compatibility.
-		 *
-		 * We must hold exec_update_mutex across this and any potential
-		 * perf_install_in_context() call for this new event to
-		 * serialize against exec() altering our credentials (and the
-		 * perf_event_exit_task() that could imply).
-		 */
-		err = -EACCES;
-		if (!perfmon_capable() && !ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS))
-			goto err_cred;
-	}
-
 	if (flags & PERF_FLAG_PID_CGROUP)
 		cgroup_fd = pid;

@@ -11745,7 +11839,7 @@ SYSCALL_DEFINE5(perf_event_open,
 				 NULL, NULL, cgroup_fd);
 	if (IS_ERR(event)) {
 		err = PTR_ERR(event);
-		goto err_cred;
+		goto err_task;
 	}

 	if (is_sampling_event(event)) {
@@ -11864,6 +11958,24 @@ SYSCALL_DEFINE5(perf_event_open,
 		goto err_context;
 	}

+	if (task) {
+		err = mutex_lock_interruptible(&task->signal->exec_update_mutex);
+		if (err)
+			goto err_file;
+
+		/*
+		 * Preserve ptrace permission check for backwards compatibility.
+		 *
+		 * We must hold exec_update_mutex across this and any potential
+		 * perf_install_in_context() call for this new event to
+		 * serialize against exec() altering our credentials (and the
+		 * perf_event_exit_task() that could imply).
+		 */
+		err = -EACCES;
+		if (!perfmon_capable() && !ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS))
+			goto err_cred;
+	}
+
 	if (move_group) {
 		gctx = __perf_event_ctx_lock_double(group_leader, ctx);

@@ -12039,7 +12151,10 @@ err_locked:
 	if (move_group)
 		perf_event_ctx_unlock(group_leader, gctx);
 	mutex_unlock(&ctx->mutex);
-/* err_file: */
+err_cred:
+	if (task)
+		mutex_unlock(&task->signal->exec_update_mutex);
+err_file:
 	fput(event_file);
 err_context:
 	perf_unpin_context(ctx);
@@ -12051,9 +12166,6 @@ err_alloc:
 	 */
 	if (!event_file)
 		free_event(event);
-err_cred:
-	if (task)
-		mutex_unlock(&task->signal->exec_update_mutex);
 err_task:
 	if (task)
 		put_task_struct(task);