[PATCH v2 4/4] uprobe: only do FOLL_SPLIT_PMD for uprobe register

[Song Liu]

Attaching uprobe to text section in THP splits the PMD mapped page table
into PTE mapped entries. On uprobe detach, we would like to regroup PMD
mapped page table entry to regain performance benefit of THP.

However, the regroup is broken For perf_event based trace_uprobe. This is
because perf_event based trace_uprobe calls uprobe_unregister twice on
close: first in TRACE_REG_PERF_CLOSE, then in TRACE_REG_PERF_UNREGISTER.
The second call will split the PMD mapped page table entry, which is not
the desired behavior.

Fix this by only use FOLL_SPLIT_PMD for uprobe register case.

Add a WARN() to confirm uprobe unregister never work on huge pages, and
abort the operation when this WARN() triggers.

Fixes: 5a52c9df62b4 ("uprobe: use FOLL_SPLIT_PMD instead of FOLL_SPLIT")
Cc: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Cc: Srikar Dronamraju <srikar@xxxxxxxxxxxxxxxxxx>
Cc: Oleg Nesterov <oleg@xxxxxxxxxx>
Signed-off-by: Song Liu <songliubraving@xxxxxx>

---
Changes v1 -> v2:
Return -EINVAL if the WARN() triggers. (Oleg Nesterov)
---
 kernel/events/uprobes.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 94d38a39d72e..c74761004ee5 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -474,14 +474,17 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, =
struct mm_struct *mm,
        struct vm_area_struct *vma;
        int ret, is_register, ref_ctr_updated =3D 0;
        bool orig_page_huge =3D false;
+       unsigned int gup_flags =3D FOLL_FORCE;

        is_register =3D is_swbp_insn(&opcode);
        uprobe =3D container_of(auprobe, struct uprobe, arch);

 retry:
+       if (is_register)
+               gup_flags |=3D FOLL_SPLIT_PMD;
        /* Read the page with vaddr into memory */
-       ret =3D get_user_pages_remote(NULL, mm, vaddr, 1,
-                       FOLL_FORCE | FOLL_SPLIT_PMD, &old_page, &vma, NULL)=
;
+       ret =3D get_user_pages_remote(NULL, mm, vaddr, 1, gup_flags,
+                                   &old_page, &vma, NULL);
        if (ret <=3D 0)
                return ret;

@@ -489,6 +492,12 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, s=
truct mm_struct *mm,
        if (ret <=3D 0)
                goto put_old;

+       if (WARN(!is_register && PageCompound(old_page),
+                "uprobe unregister should never work on compound page\n"))=
 {
+               ret =3D -EINVAL;
+               goto put_old;
+       }
+
        /* We are going to replace instruction, update ref_ctr. */
        if (!ref_ctr_updated && uprobe->ref_ctr_offset) {
                ret =3D update_ref_ctr(uprobe, mm, is_register ? 1 : -1);
--
2.17.1