From: Ding Tianhong <dingtianhong@huawei.com> ascend inclusion category: feature bugzilla: NA CVE: NA ------------------------------------------------- The fork() will create the new mm for new process, the mm should not take any information from the parent process, so need to clean it. The exit() will mmput the mm and free the memory, if the mm is alrready be used for sp_group, need to clean the group first. Signed-off-by: Tang Yizhou <tangyizhou@huawei.com> Signed-off-by: Li Ming <limingming.li@huawei.com> Signed-off-by: Zefan Li <lizefan@huawei.com> Signed-off-by: Zhou Guanghui <zhouguanghui1@huawei.com> Signed-off-by: Ding Tianhong <dingtianhong@huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang@huawei.com> Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> --- kernel/fork.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/kernel/fork.c b/kernel/fork.c index 1ac49d1852cf..b5f9a36fa4eb 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -91,6 +91,7 @@ #include <linux/kcov.h> #include <linux/livepatch.h> #include <linux/thread_info.h> +#include <linux/share_pool.h> #include <asm/pgtable.h> #include <asm/pgalloc.h> @@ -1023,6 +1024,9 @@ static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p, goto fail_nocontext; mm->user_ns = get_user_ns(user_ns); + + sp_init_mm(mm); + return mm; fail_nocontext: @@ -1051,11 +1055,16 @@ static inline void __mmput(struct mm_struct *mm) { VM_BUG_ON(atomic_read(&mm->mm_users)); + sp_group_exit(mm); + uprobe_clear_state(mm); exit_aio(mm); ksm_exit(mm); khugepaged_exit(mm); /* must run before exit_mmap */ exit_mmap(mm); + + sp_group_post_exit(mm); + mm_put_huge_zero_page(mm); set_mm_exe_file(mm, NULL); if (!list_empty(&mm->mmlist)) { -- 2.25.1