From b56e64466554dfd39a7695f5025f583d8a20288d Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Fri, 21 Mar 2025 12:37:11 +0100 Subject: [PATCH] kernel/events/uprobes: pass VMA instead of MM to remove_breakpoint() Patch series "kernel/events/uprobes: uprobe_write_opcode() rewrite", v3. Currently, uprobe_write_opcode() implements COW-breaking manually, which is really far from ideal. Further, there is interest in supporting uprobes on hugetlb pages [1], and leaving at least the COW-breaking to the core will make this much easier. Also, I think the current code doesn't really handle some things properly (see patch #3) when replacing/zapping pages. Let's rewrite it, to leave COW-breaking to the fault handler, and handle registration/unregistration by temporarily unmapping the anonymous page, modifying it, and mapping it again. We still have to implement zapping of anonymous pages ourselves, unfortunately. We could look into not performing the temporary unmapping if we can perform the write atomically, which would likely also make adding hugetlb support a lot easier. But, limited (e.g., only PMD/PUD) hugetlb support could be added on top of this with some tweaking. Note that we now won't have to allocate another anonymous folio when unregistering (which will be beneficial for hugetlb as well), we can simply modify the already-mapped one from the registration (if any). When registering a uprobe, we'll first trigger a ptrace-like write fault to break COW, to then modify the already-mapped page. Briefly sanity tested with perf probes and with the bpf uprobes selftest. This patch (of 3): Pass VMA instead of MM to remove_breakpoint() and remove the "MM" argument from install_breakpoint(), because it can easily be derived from the VMA. Link: https://lkml.kernel.org/r/20250321113713.204682-1-david@redhat.com Link: https://lkml.kernel.org/r/20250321113713.204682-2-david@redhat.com Signed-off-by: David Hildenbrand Acked-by: Oleg Nesterov Acked-by: Peter Zijlstra (Intel) Cc: Jiri Olsa Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Andrii Nakryiko Cc: Arnaldo Carvalho de Melo Cc: Ian Rogers Cc: Ingo Molnar Cc: Kan Liang Cc: Mark Rutland Cc: "Masami Hiramatsu (Google)" Cc: Matthew Wilcox (Oracle) Cc: Namhyung kim Cc: Russel King Cc: tongtiangen Signed-off-by: Andrew Morton --- kernel/events/uprobes.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c index 8d783b5882b6..8dc23ca9f66f 100644 --- a/kernel/events/uprobes.c +++ b/kernel/events/uprobes.c @@ -1134,10 +1134,10 @@ static bool filter_chain(struct uprobe *uprobe, struct mm_struct *mm) return ret; } -static int -install_breakpoint(struct uprobe *uprobe, struct mm_struct *mm, - struct vm_area_struct *vma, unsigned long vaddr) +static int install_breakpoint(struct uprobe *uprobe, struct vm_area_struct *vma, + unsigned long vaddr) { + struct mm_struct *mm = vma->vm_mm; bool first_uprobe; int ret; @@ -1162,9 +1162,11 @@ install_breakpoint(struct uprobe *uprobe, struct mm_struct *mm, return ret; } -static int -remove_breakpoint(struct uprobe *uprobe, struct mm_struct *mm, unsigned long vaddr) +static int remove_breakpoint(struct uprobe *uprobe, struct vm_area_struct *vma, + unsigned long vaddr) { + struct mm_struct *mm = vma->vm_mm; + set_bit(MMF_RECALC_UPROBES, &mm->flags); return set_orig_insn(&uprobe->arch, mm, vaddr); } @@ -1296,10 +1298,10 @@ register_for_each_vma(struct uprobe *uprobe, struct uprobe_consumer *new) if (is_register) { /* consult only the "caller", new consumer. */ if (consumer_filter(new, mm)) - err = install_breakpoint(uprobe, mm, vma, info->vaddr); + err = install_breakpoint(uprobe, vma, info->vaddr); } else if (test_bit(MMF_HAS_UPROBES, &mm->flags)) { if (!filter_chain(uprobe, mm)) - err |= remove_breakpoint(uprobe, mm, info->vaddr); + err |= remove_breakpoint(uprobe, vma, info->vaddr); } unlock: @@ -1472,7 +1474,7 @@ static int unapply_uprobe(struct uprobe *uprobe, struct mm_struct *mm) continue; vaddr = offset_to_vaddr(vma, uprobe->offset); - err |= remove_breakpoint(uprobe, mm, vaddr); + err |= remove_breakpoint(uprobe, vma, vaddr); } mmap_read_unlock(mm); @@ -1610,7 +1612,7 @@ int uprobe_mmap(struct vm_area_struct *vma) if (!fatal_signal_pending(current) && filter_chain(uprobe, vma->vm_mm)) { unsigned long vaddr = offset_to_vaddr(vma, uprobe->offset); - install_breakpoint(uprobe, vma->vm_mm, vma, vaddr); + install_breakpoint(uprobe, vma, vaddr); } put_uprobe(uprobe); } -- 2.25.1