The patch titled
Subject: bpf: rework the compat kernel probe handling
has been added to the -mm tree. Its filename is
bpf-rework-the-compat-kernel-probe-handling.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/bpf-rework-the-compat-kernel-probe-handling.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/bpf-rework-the-compat-kernel-probe-handling.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Christoph Hellwig <hch@xxxxxx>
Subject: bpf: rework the compat kernel probe handling
Instead of using the dangerous probe_kernel_read and strncpy_from_unsafe
helpers, rework the compat probes to check if an address is a kernel or
userspace one, and then use the low-level kernel or user probe helper
shared by the proper kernel and user probe helpers. This slightly
changes behavior as the compat probe on a user address doesn't check
the lockdown flags, just as the pure user probes do.
Link: http://lkml.kernel.org/r/20200521152301.2587579-14-hch@xxxxxx
Signed-off-by: Christoph Hellwig <hch@xxxxxx>
Cc: Alexei Starovoitov <ast@xxxxxxxxxx>
Cc: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxx>
Cc: Masami Hiramatsu <mhiramat@xxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
kernel/trace/bpf_trace.c | 109 ++++++++++++++++++++++---------------
1 file changed, 67 insertions(+), 42 deletions(-)
--- a/kernel/trace/bpf_trace.c~bpf-rework-the-compat-kernel-probe-handling
+++ a/kernel/trace/bpf_trace.c
@@ -136,17 +136,23 @@ static const struct bpf_func_proto bpf_o
};
#endif
-BPF_CALL_3(bpf_probe_read_user, void *, dst, u32, size,
- const void __user *, unsafe_ptr)
+static __always_inline int
+bpf_probe_read_user_common(void *dst, u32 size, const void __user *unsafe_ptr)
{
- int ret = probe_user_read(dst, unsafe_ptr, size);
+ int ret;
+ ret = probe_user_read(dst, unsafe_ptr, size);
if (unlikely(ret < 0))
memset(dst, 0, size);
-
return ret;
}
+BPF_CALL_3(bpf_probe_read_user, void *, dst, u32, size,
+ const void __user *, unsafe_ptr)
+{
+ return bpf_probe_read_user_common(dst, size, unsafe_ptr);
+}
+
static const struct bpf_func_proto bpf_probe_read_user_proto = {
.func = bpf_probe_read_user,
.gpl_only = true,
@@ -156,17 +162,24 @@ static const struct bpf_func_proto bpf_p
.arg3_type = ARG_ANYTHING,
};
-BPF_CALL_3(bpf_probe_read_user_str, void *, dst, u32, size,
- const void __user *, unsafe_ptr)
+static __always_inline int
+bpf_probe_read_user_str_common(void *dst, u32 size,
+ const void __user *unsafe_ptr)
{
- int ret = strncpy_from_user_nofault(dst, unsafe_ptr, size);
+ int ret;
+ ret = strncpy_from_user_nofault(dst, unsafe_ptr, size);
if (unlikely(ret < 0))
memset(dst, 0, size);
-
return ret;
}
+BPF_CALL_3(bpf_probe_read_user_str, void *, dst, u32, size,
+ const void __user *, unsafe_ptr)
+{
+ return bpf_probe_read_user_str_common(dst, size, unsafe_ptr);
+}
+
static const struct bpf_func_proto bpf_probe_read_user_str_proto = {
.func = bpf_probe_read_user_str,
.gpl_only = true,
@@ -177,25 +190,25 @@ static const struct bpf_func_proto bpf_p
};
static __always_inline int
-bpf_probe_read_kernel_common(void *dst, u32 size, const void *unsafe_ptr,
- const bool compat)
+bpf_probe_read_kernel_common(void *dst, u32 size, const void *unsafe_ptr)
{
int ret = security_locked_down(LOCKDOWN_BPF_READ);
if (unlikely(ret < 0))
- goto out;
- ret = compat ? probe_kernel_read(dst, unsafe_ptr, size) :
- probe_kernel_read_strict(dst, unsafe_ptr, size);
+ goto fail;
+ ret = probe_kernel_read_strict(dst, unsafe_ptr, size);
if (unlikely(ret < 0))
-out:
- memset(dst, 0, size);
+ goto fail;
+ return ret;
+fail:
+ memset(dst, 0, size);
return ret;
}
BPF_CALL_3(bpf_probe_read_kernel, void *, dst, u32, size,
const void *, unsafe_ptr)
{
- return bpf_probe_read_kernel_common(dst, size, unsafe_ptr, false);
+ return bpf_probe_read_kernel_common(dst, size, unsafe_ptr);
}
static const struct bpf_func_proto bpf_probe_read_kernel_proto = {
@@ -207,50 +220,37 @@ static const struct bpf_func_proto bpf_p
.arg3_type = ARG_ANYTHING,
};
-BPF_CALL_3(bpf_probe_read_compat, void *, dst, u32, size,
- const void *, unsafe_ptr)
-{
- return bpf_probe_read_kernel_common(dst, size, unsafe_ptr, true);
-}
-
-static const struct bpf_func_proto bpf_probe_read_compat_proto = {
- .func = bpf_probe_read_compat,
- .gpl_only = true,
- .ret_type = RET_INTEGER,
- .arg1_type = ARG_PTR_TO_UNINIT_MEM,
- .arg2_type = ARG_CONST_SIZE_OR_ZERO,
- .arg3_type = ARG_ANYTHING,
-};
-
static __always_inline int
-bpf_probe_read_kernel_str_common(void *dst, u32 size, const void *unsafe_ptr,
- const bool compat)
+bpf_probe_read_kernel_str_common(void *dst, u32 size, const void *unsafe_ptr)
{
int ret = security_locked_down(LOCKDOWN_BPF_READ);
if (unlikely(ret < 0))
- goto out;
+ goto fail;
+
/*
- * The strncpy_from_unsafe_*() call will likely not fill the entire
- * buffer, but that's okay in this circumstance as we're probing
+ * The strncpy_from_kernel_nofault() call will likely not fill the
+ * entire buffer, but that's okay in this circumstance as we're probing
* arbitrary memory anyway similar to bpf_probe_read_*() and might
* as well probe the stack. Thus, memory is explicitly cleared
* only in error case, so that improper users ignoring return
* code altogether don't copy garbage; otherwise length of string
* is returned that can be used for bpf_perf_event_output() et al.
*/
- ret = compat ? strncpy_from_unsafe(dst, unsafe_ptr, size) :
- strncpy_from_kernel_nofault(dst, unsafe_ptr, size);
+ ret = strncpy_from_kernel_nofault(dst, unsafe_ptr, size);
if (unlikely(ret < 0))
-out:
- memset(dst, 0, size);
+ goto fail;
+
+ return 0;
+fail:
+ memset(dst, 0, size);
return ret;
}
BPF_CALL_3(bpf_probe_read_kernel_str, void *, dst, u32, size,
const void *, unsafe_ptr)
{
- return bpf_probe_read_kernel_str_common(dst, size, unsafe_ptr, false);
+ return bpf_probe_read_kernel_str_common(dst, size, unsafe_ptr);
}
static const struct bpf_func_proto bpf_probe_read_kernel_str_proto = {
@@ -262,10 +262,34 @@ static const struct bpf_func_proto bpf_p
.arg3_type = ARG_ANYTHING,
};
+#ifdef CONFIG_ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
+BPF_CALL_3(bpf_probe_read_compat, void *, dst, u32, size,
+ const void *, unsafe_ptr)
+{
+ if ((unsigned long)unsafe_ptr < TASK_SIZE) {
+ return bpf_probe_read_user_common(dst, size,
+ (__force void __user *)unsafe_ptr);
+ }
+ return bpf_probe_read_kernel_common(dst, size, unsafe_ptr);
+}
+
+static const struct bpf_func_proto bpf_probe_read_compat_proto = {
+ .func = bpf_probe_read_compat,
+ .gpl_only = true,
+ .ret_type = RET_INTEGER,
+ .arg1_type = ARG_PTR_TO_UNINIT_MEM,
+ .arg2_type = ARG_CONST_SIZE_OR_ZERO,
+ .arg3_type = ARG_ANYTHING,
+};
+
BPF_CALL_3(bpf_probe_read_compat_str, void *, dst, u32, size,
const void *, unsafe_ptr)
{
- return bpf_probe_read_kernel_str_common(dst, size, unsafe_ptr, true);
+ if ((unsigned long)unsafe_ptr < TASK_SIZE) {
+ return bpf_probe_read_user_str_common(dst, size,
+ (__force void __user *)unsafe_ptr);
+ }
+ return bpf_probe_read_kernel_str_common(dst, size, unsafe_ptr);
}
static const struct bpf_func_proto bpf_probe_read_compat_str_proto = {
@@ -276,6 +300,7 @@ static const struct bpf_func_proto bpf_p
.arg2_type = ARG_CONST_SIZE_OR_ZERO,
.arg3_type = ARG_ANYTHING,
};
+#endif /* CONFIG_ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE */
BPF_CALL_3(bpf_probe_write_user, void __user *, unsafe_ptr, const void *, src,
u32, size)
_
Patches currently in -mm which might be from hch@xxxxxx are
x86-hyperv-use-vmalloc_exec-for-the-hypercall-page.patch
x86-fix-vmap-arguments-in-map_irq_stack.patch
staging-android-ion-use-vmap-instead-of-vm_map_ram.patch
staging-media-ipu3-use-vmap-instead-of-reimplementing-it.patch
dma-mapping-use-vmap-insted-of-reimplementing-it.patch
powerpc-add-an-ioremap_phb-helper.patch
powerpc-remove-__ioremap_at-and-__iounmap_at.patch
mm-remove-__get_vm_area.patch
mm-unexport-unmap_kernel_range_noflush.patch
mm-rename-config_pgtable_mapping-to-config_zsmalloc_pgtable_mapping.patch
mm-only-allow-page-table-mappings-for-built-in-zsmalloc.patch
mm-pass-addr-as-unsigned-long-to-vb_free.patch
mm-remove-vmap_page_range_noflush-and-vunmap_page_range.patch
mm-rename-vmap_page_range-to-map_kernel_range.patch
mm-dont-return-the-number-of-pages-from-map_kernel_range_noflush.patch
mm-remove-map_vm_range.patch
mm-remove-unmap_vmap_area.patch
mm-remove-the-prot-argument-from-vm_map_ram.patch
mm-enforce-that-vmap-cant-map-pages-executable.patch
gpu-drm-remove-the-powerpc-hack-in-drm_legacy_sg_alloc.patch
mm-remove-the-pgprot-argument-to-__vmalloc.patch
mm-remove-the-prot-argument-to-__vmalloc_node.patch
mm-remove-both-instances-of-__vmalloc_node_flags.patch
mm-remove-__vmalloc_node_flags_caller.patch
mm-switch-the-test_vmalloc-module-to-use-__vmalloc_node.patch
mm-remove-vmalloc_user_node_flags.patch
arm64-use-__vmalloc_node-in-arch_alloc_vmap_stack.patch
powerpc-use-__vmalloc_node-in-alloc_vm_stack.patch
s390-use-__vmalloc_node-in-stack_alloc.patch
exec-simplify-the-copy_strings_kernel-calling-convention.patch
exec-open-code-copy_string_kernel.patch
amdgpu-a-null-mm-does-not-mean-a-thread-is-a-kthread.patch
kernel-move-use_mm-unuse_mm-to-kthreadc.patch
kernel-move-use_mm-unuse_mm-to-kthreadc-v2.patch
kernel-better-document-the-use_mm-unuse_mm-api-contract.patch
kernel-better-document-the-use_mm-unuse_mm-api-contract-v2.patch
kernel-set-user_ds-in-kthread_use_mm.patch
arm-fix-the-flush_icache_range-arguments-in-set_fiq_handler.patch
nds32-unexport-flush_icache_page.patch
powerpc-unexport-flush_icache_user_range.patch
unicore32-remove-flush_cache_user_range.patch
asm-generic-fix-the-inclusion-guards-for-cacheflushh.patch
asm-generic-dont-include-linux-mmh-in-cacheflushh.patch
asm-generic-dont-include-linux-mmh-in-cacheflushh-fix.patch
asm-generic-improve-the-flush_dcache_page-stub.patch
alpha-use-asm-generic-cacheflushh.patch
arm64-use-asm-generic-cacheflushh.patch
c6x-use-asm-generic-cacheflushh.patch
hexagon-use-asm-generic-cacheflushh.patch
ia64-use-asm-generic-cacheflushh.patch
microblaze-use-asm-generic-cacheflushh.patch
m68knommu-use-asm-generic-cacheflushh.patch
openrisc-use-asm-generic-cacheflushh.patch
powerpc-use-asm-generic-cacheflushh.patch
riscv-use-asm-generic-cacheflushh.patch
armsparcunicore32-remove-flush_icache_user_range.patch
mm-rename-flush_icache_user_range-to-flush_icache_user_page.patch
asm-generic-add-a-flush_icache_user_range-stub.patch
sh-implement-flush_icache_user_range.patch
xtensa-implement-flush_icache_user_range.patch
arm-rename-flush_cache_user_range-to-flush_icache_user_range.patch
m68k-implement-flush_icache_user_range.patch
exec-only-build-read_code-when-needed.patch
exec-use-flush_icache_user_range-in-read_code.patch
binfmt_flat-use-flush_icache_user_range.patch
nommu-use-flush_icache_user_range-in-brk-and-mmap.patch
module-move-the-set_fs-hack-for-flush_icache_range-to-m68k.patch
maccess-unexport-probe_kernel_write-and-probe_user_write.patch
maccess-remove-various-unused-weak-aliases.patch
maccess-remove-duplicate-kerneldoc-comments.patch
maccess-clarify-kerneldoc-comments.patch
maccess-update-the-top-of-file-comment.patch
maccess-rename-strncpy_from_unsafe_user-to-strncpy_from_user_nofault.patch
maccess-rename-strncpy_from_unsafe_strict-to-strncpy_from_kernel_nofault.patch
maccess-rename-strnlen_unsafe_user-to-strnlen_user_nofault.patch
maccess-remove-probe_read_common-and-probe_write_common.patch
maccess-unify-the-probe-kernel-arch-hooks.patch
bpf-factor-out-a-bpf_trace_copy_string-helper.patch
bpf-handle-the-compat-string-in-bpf_trace_copy_string-better.patch
bpf-rework-the-compat-kernel-probe-handling.patch
tracing-kprobes-handle-mixed-kernel-userspace-probes-better.patch
maccess-remove-strncpy_from_unsafe.patch
maccess-always-use-strict-semantics-for-probe_kernel_read.patch
maccess-move-user-access-routines-together.patch
maccess-allow-architectures-to-provide-kernel-probing-directly.patch
x86-use-non-set_fs-based-maccess-routines.patch
maccess-return-erange-when-copy_from_kernel_nofault_allowed-fails.patch
