This adds a Kconfig option and boot param to allow removing
the FOLL_FORCE flag from /proc/pid/mem write calls because
it can be abused.

The traditional forcing behavior is kept as default because
it can break GDB and some other use cases.

Previously we tried a more sophisticated approach allowing
distributions to fine-tune /proc/pid/mem behavior, however
that got NAK-ed by Linus [1], who prefers this simpler
approach with semantics also easier to understand for users.

Link: https://lore.kernel.org/lkml/CAHk-=wiGWLChxYmUA5HrT5aopZrB7_2VTa0NLZcxORgkUe5tEQ@mail.gmail.com/ (local) [1]
Cc: Doug Anderson <dianders@chromium.org>
Cc: Jeff Xu <redacted>
Cc: Jann Horn <jannh@google.com>
Cc: Kees Cook <kees@kernel.org>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Christian Brauner <brauner@kernel.org>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Adrian Ratiu <redacted>
---
Changes in v4:
* Fixed doc punctuation, used passive tense, improved
  wording consistency, fixed default value wording
* Made struct constant_table a static const __initconst
* Reworked proc_mem_foll_force() indentation and var
  declarations to make code clearer
* Reworked enum + struct definition so lookup_constant()
  defaults to 'always'.

Changes in v3:
* Simplified code to use shorthand ifs and a
  lookup_constant() table

Changes in v2:
* Added bootparam on top of Linus' patch
* Slightly reworded commit msg
---
 .../admin-guide/kernel-parameters.txt         | 10 ++++
 fs/proc/base.c                                | 54 ++++++++++++++++++-
 security/Kconfig                              | 32 +++++++++++
 3 files changed, 95 insertions(+), 1 deletion(-)

diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index f1384c7b59c9..8396e015aab3 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt

@@ -4788,6 +4788,16 @@
        printk.time=    Show timing data prefixed to each printk message line
                        Format: <bool>  (1/Y/y=enable, 0/N/n=disable)

+       proc_mem.force_override= [KNL]
+                       Format: {always | ptrace | never}
+                       Traditionally /proc/pid/mem allows memory permissions to be
+                       overridden without restrictions. This option may be set to
+                       restrict that. Can be one of:
+                       - 'always': traditional behavior always allows mem overrides.
+                       - 'ptrace': only allow mem overrides for active ptracers.
+                       - 'never':  never allow mem overrides.
+                       If not specified, default is the CONFIG_PROC_MEM_* choice.
+
        processor.max_cstate=   [HW,ACPI]
                        Limit processor to maximum C-state
                        max_cstate=9 overrides any DMI blacklist limit.

diff --git a/fs/proc/base.c b/fs/proc/base.c
index 72a1acd03675..daacb8070042 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c

@@ -85,6 +85,7 @@
 #include <linux/elf.h>
 #include <linux/pid_namespace.h>
 #include <linux/user_namespace.h>
+#include <linux/fs_parser.h>
 #include <linux/fs_struct.h>
 #include <linux/slab.h>
 #include <linux/sched/autogroup.h>

@@ -117,6 +118,35 @@
 static u8 nlink_tid __ro_after_init;
 static u8 nlink_tgid __ro_after_init;

+enum proc_mem_force {
+       PROC_MEM_FORCE_ALWAYS,
+       PROC_MEM_FORCE_PTRACE,
+       PROC_MEM_FORCE_NEVER
+};
+
+static enum proc_mem_force proc_mem_force_override __ro_after_init =
+       IS_ENABLED(CONFIG_PROC_MEM_NO_FORCE) ? PROC_MEM_FORCE_NEVER :
+       IS_ENABLED(CONFIG_PROC_MEM_FORCE_PTRACE) ? PROC_MEM_FORCE_PTRACE :
+       PROC_MEM_FORCE_ALWAYS;
+
+static const struct constant_table proc_mem_force_table[] __initconst = {
+       { "never", PROC_MEM_FORCE_NEVER },
+       { "ptrace", PROC_MEM_FORCE_PTRACE },
+       { }
+};
+
+static int __init early_proc_mem_force_override(char *buf)
+{
+       if (!buf)
+               return -EINVAL;
+
+       proc_mem_force_override = lookup_constant(proc_mem_force_table,
+                                                 buf, PROC_MEM_FORCE_ALWAYS);

+
+       return 0;
+}
+early_param("proc_mem.force_override", early_proc_mem_force_override);
+
 struct pid_entry {
        const char *name;
        unsigned int len;

@@ -835,6 +865,26 @@ static int mem_open(struct inode *inode, struct file *file)
        return ret;
 }

+static bool proc_mem_foll_force(struct file *file, struct mm_struct *mm)
+{
+       struct task_struct *task;
+       bool ptrace_active = false;
+
+       switch (proc_mem_force_override) {
+       case PROC_MEM_FORCE_NEVER:
+               return false;
+       case PROC_MEM_FORCE_PTRACE:
+               task = get_proc_task(file_inode(file));
+               if (task) {
+                       ptrace_active = task->ptrace && task->mm == mm && task->parent == current;

+                       put_task_struct(task);
+               }
+               return ptrace_active;
+       default:
+               return true;
+       }
+}
+
 static ssize_t mem_rw(struct file *file, char __user *buf,
                        size_t count, loff_t *ppos, int write)
 {

@@ -855,7 +905,9 @@ static ssize_t mem_rw(struct file *file, char __user *buf,
        if (!mmget_not_zero(mm))
                goto free;

-       flags = FOLL_FORCE | (write ? FOLL_WRITE : 0);
+       flags = write ? FOLL_WRITE : 0;
+       if (proc_mem_foll_force(file, mm))
+               flags |= FOLL_FORCE;

        while (count > 0) {
                size_t this_len = min_t(size_t, count, PAGE_SIZE);

diff --git a/security/Kconfig b/security/Kconfig
index 412e76f1575d..a93c1a9b7c28 100644
--- a/security/Kconfig
+++ b/security/Kconfig

@@ -19,6 +19,38 @@ config SECURITY_DMESG_RESTRICT

          If you are unsure how to answer this question, answer N.

+choice
+       prompt "Allow /proc/pid/mem access override"
+       default PROC_MEM_ALWAYS_FORCE
+       help
+         Traditionally /proc/pid/mem allows users to override memory
+         permissions for users like ptrace, assuming they have ptrace
+         capability.
+
+         This allows people to limit that - either never override, or
+         require actual active ptrace attachment.
+
+         Defaults to the traditional behavior (for now)
+
+config PROC_MEM_ALWAYS_FORCE
+       bool "Traditional /proc/pid/mem behavior"
+       help
+         This allows /proc/pid/mem accesses to override memory mapping
+         permissions if you have ptrace access rights.
+
+config PROC_MEM_FORCE_PTRACE
+       bool "Require active ptrace() use for access override"
+       help
+         This allows /proc/pid/mem accesses to override memory mapping
+         permissions for active ptracers like gdb.
+
+config PROC_MEM_NO_FORCE
+       bool "Never"
+       help
+         Never override memory mapping permissions
+
+endchoice
+
 config SECURITY
        bool "Enable different security models"
        depends on SYSFS
--

2.44.2