Message ID | 20240329094906.18147-3-ubizjak@gmail.com (mailing list archive) |
---|---|
State | Superseded |
Delegated to: | BPF |
Headers | show |
Series | x86/bpf: Fixes for the BPF JIT with retbleed=stuff | expand |
On Fri, Mar 29, 2024 at 2:49 AM Uros Bizjak <ubizjak@gmail.com> wrote: > > From: Joan Bruguera Micó <joanbrugueram@gmail.com> > > The recently introduced support for %rip-relative relocations in the > call thunk template assumes that the code is being patched in-place, > so the destination of the relocation matches the address of the code. > This is not true for the call depth accounting emitted by the BPF JIT, > so the calculated address is wrong and usually causes a page fault. Could you share the link to what this 'rip-relative' relocation is ? > Pass the destination IP when the BPF JIT emits call depth accounting. > > Fixes: 17bce3b2ae2d ("x86/callthunks: Handle %rip-relative relocations in call thunk template") Ohh. It's buried inside that patch. Pls make commit log a bit more clear that that commit 17bce3b2ae2d broke x86_call_depth_emit_accounting logic. > Signed-off-by: Joan Bruguera Micó <joanbrugueram@gmail.com> > Reviewed-by: Uros Bizjak <ubizjak@gmail.com> > Acked-by: Ingo Molnar <mingo@kernel.org> > Cc: Alexei Starovoitov <ast@kernel.org> > Cc: Daniel Borkmann <daniel@iogearbox.net> > --- > arch/x86/include/asm/alternative.h | 4 ++-- > arch/x86/kernel/callthunks.c | 4 ++-- > arch/x86/net/bpf_jit_comp.c | 19 ++++++++----------- > 3 files changed, 12 insertions(+), 15 deletions(-) > > diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h > index fcd20c6dc7f9..67b68d0d17d1 100644 > --- a/arch/x86/include/asm/alternative.h > +++ b/arch/x86/include/asm/alternative.h > @@ -117,7 +117,7 @@ extern void callthunks_patch_builtin_calls(void); > extern void callthunks_patch_module_calls(struct callthunk_sites *sites, > struct module *mod); > extern void *callthunks_translate_call_dest(void *dest); > -extern int x86_call_depth_emit_accounting(u8 **pprog, void *func); > +extern int x86_call_depth_emit_accounting(u8 **pprog, void *func, void *ip); > #else > static __always_inline void callthunks_patch_builtin_calls(void) {} > static __always_inline void > @@ -128,7 +128,7 @@ static __always_inline void *callthunks_translate_call_dest(void *dest) > return dest; > } > static __always_inline int x86_call_depth_emit_accounting(u8 **pprog, > - void *func) > + void *func, void *ip) > { > return 0; > } > diff --git a/arch/x86/kernel/callthunks.c b/arch/x86/kernel/callthunks.c > index 30335182b6b0..e92ff0c11db8 100644 > --- a/arch/x86/kernel/callthunks.c > +++ b/arch/x86/kernel/callthunks.c > @@ -314,7 +314,7 @@ static bool is_callthunk(void *addr) > return !bcmp(pad, insn_buff, tmpl_size); > } > > -int x86_call_depth_emit_accounting(u8 **pprog, void *func) > +int x86_call_depth_emit_accounting(u8 **pprog, void *func, void *ip) > { > unsigned int tmpl_size = SKL_TMPL_SIZE; > u8 insn_buff[MAX_PATCH_LEN]; > @@ -327,7 +327,7 @@ int x86_call_depth_emit_accounting(u8 **pprog, void *func) > return 0; > > memcpy(insn_buff, skl_call_thunk_template, tmpl_size); > - apply_relocation(insn_buff, tmpl_size, *pprog, > + apply_relocation(insn_buff, tmpl_size, ip, Did the logic inside apply_relocation() change to have a new meaning for 'dest' and 'src'? Answering to myself... yes. in that commit. Better commit log would have made the code review so much easier. > skl_call_thunk_template, tmpl_size); > > memcpy(*pprog, insn_buff, tmpl_size); > diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c > index 09f7dc9d4d65..f2e8769f5eee 100644 > --- a/arch/x86/net/bpf_jit_comp.c > +++ b/arch/x86/net/bpf_jit_comp.c > @@ -481,7 +481,7 @@ static int emit_rsb_call(u8 **pprog, void *func, void *ip) > { > void *adjusted_ip; > OPTIMIZER_HIDE_VAR(func); > - adjusted_ip = ip + x86_call_depth_emit_accounting(pprog, func); > + adjusted_ip = ip + x86_call_depth_emit_accounting(pprog, func, ip); Now I see why you added extra var in the previous patch. Should have mentioned it in the commit log. > return emit_patch(pprog, func, adjusted_ip, 0xE8); > } > > @@ -1973,20 +1973,17 @@ st: if (is_imm8(insn->off)) > > /* call */ > case BPF_JMP | BPF_CALL: { > - int offs; > + u8 *ip = image + addrs[i - 1]; > > func = (u8 *) __bpf_call_base + imm32; > if (tail_call_reachable) { > RESTORE_TAIL_CALL_CNT(bpf_prog->aux->stack_depth); > - if (!imm32) > - return -EINVAL; > - offs = 7 + x86_call_depth_emit_accounting(&prog, func); > - } else { > - if (!imm32) > - return -EINVAL; > - offs = x86_call_depth_emit_accounting(&prog, func); > + ip += 7; > } > - if (emit_call(&prog, func, image + addrs[i - 1] + offs)) > + if (!imm32) > + return -EINVAL; > + ip += x86_call_depth_emit_accounting(&prog, func, ip); > + if (emit_call(&prog, func, ip)) > return -EINVAL; > break; > } > @@ -2836,7 +2833,7 @@ static int __arch_prepare_bpf_trampoline(struct bpf_tramp_image *im, void *rw_im > * Direct-call fentry stub, as such it needs accounting for the > * __fentry__ call. > */ > - x86_call_depth_emit_accounting(&prog, NULL); > + x86_call_depth_emit_accounting(&prog, NULL, image); Overall it all makes sense. Pls respin with more precise commit logs.
On Fri, Mar 29, 2024 at 10:53 PM Alexei Starovoitov <alexei.starovoitov@gmail.com> wrote: > > On Fri, Mar 29, 2024 at 2:49 AM Uros Bizjak <ubizjak@gmail.com> wrote: > > > > From: Joan Bruguera Micó <joanbrugueram@gmail.com> > > > > The recently introduced support for %rip-relative relocations in the > > call thunk template assumes that the code is being patched in-place, > > so the destination of the relocation matches the address of the code. > > This is not true for the call depth accounting emitted by the BPF JIT, > > so the calculated address is wrong and usually causes a page fault. > > Could you share the link to what this 'rip-relative' relocation is ? Please see the "RIP relative addressing" section in [1]. [1] https://compas.cs.stonybrook.edu/~nhonarmand/courses/sp17/cse506/ref/assembly.html In our case: The callthunks patching creates a call thunk template in the .rodata section (please see arch/x86/kernel/callthunks.c) that is later copied to the .text section at the correct place. The template uses X86_call_depth in the pcpu_hot structure. Previously, the template used absolute location for X86_call_depth and the linker resolved the address in the template to this absolute location. There is no issue when this template is copied to the various places in the .text section. When we want to use PC relative relocations (to reduce the code size), then the linker calculates the address of the variable in the template according to the PC in the .rodata section. If we want to copy the template to its final location, then the address of X86_call_depth, relative to the PC, has to be adjusted, as explained in arch/x86/kernel/alternative.c, in the comment above apply_reloc_n macro. Uros. > > Pass the destination IP when the BPF JIT emits call depth accounting. > > > > Fixes: 17bce3b2ae2d ("x86/callthunks: Handle %rip-relative relocations in call thunk template") > > Ohh. It's buried inside that patch. > Pls make commit log a bit more clear that that commit 17bce3b2ae2d > broke x86_call_depth_emit_accounting logic. > > > Signed-off-by: Joan Bruguera Micó <joanbrugueram@gmail.com> > > Reviewed-by: Uros Bizjak <ubizjak@gmail.com> > > Acked-by: Ingo Molnar <mingo@kernel.org> > > Cc: Alexei Starovoitov <ast@kernel.org> > > Cc: Daniel Borkmann <daniel@iogearbox.net> > > --- > > arch/x86/include/asm/alternative.h | 4 ++-- > > arch/x86/kernel/callthunks.c | 4 ++-- > > arch/x86/net/bpf_jit_comp.c | 19 ++++++++----------- > > 3 files changed, 12 insertions(+), 15 deletions(-) > > > > diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h > > index fcd20c6dc7f9..67b68d0d17d1 100644 > > --- a/arch/x86/include/asm/alternative.h > > +++ b/arch/x86/include/asm/alternative.h > > @@ -117,7 +117,7 @@ extern void callthunks_patch_builtin_calls(void); > > extern void callthunks_patch_module_calls(struct callthunk_sites *sites, > > struct module *mod); > > extern void *callthunks_translate_call_dest(void *dest); > > -extern int x86_call_depth_emit_accounting(u8 **pprog, void *func); > > +extern int x86_call_depth_emit_accounting(u8 **pprog, void *func, void *ip); > > #else > > static __always_inline void callthunks_patch_builtin_calls(void) {} > > static __always_inline void > > @@ -128,7 +128,7 @@ static __always_inline void *callthunks_translate_call_dest(void *dest) > > return dest; > > } > > static __always_inline int x86_call_depth_emit_accounting(u8 **pprog, > > - void *func) > > + void *func, void *ip) > > { > > return 0; > > } > > diff --git a/arch/x86/kernel/callthunks.c b/arch/x86/kernel/callthunks.c > > index 30335182b6b0..e92ff0c11db8 100644 > > --- a/arch/x86/kernel/callthunks.c > > +++ b/arch/x86/kernel/callthunks.c > > @@ -314,7 +314,7 @@ static bool is_callthunk(void *addr) > > return !bcmp(pad, insn_buff, tmpl_size); > > } > > > > -int x86_call_depth_emit_accounting(u8 **pprog, void *func) > > +int x86_call_depth_emit_accounting(u8 **pprog, void *func, void *ip) > > { > > unsigned int tmpl_size = SKL_TMPL_SIZE; > > u8 insn_buff[MAX_PATCH_LEN]; > > @@ -327,7 +327,7 @@ int x86_call_depth_emit_accounting(u8 **pprog, void *func) > > return 0; > > > > memcpy(insn_buff, skl_call_thunk_template, tmpl_size); > > - apply_relocation(insn_buff, tmpl_size, *pprog, > > + apply_relocation(insn_buff, tmpl_size, ip, > > Did the logic inside apply_relocation() change to have > a new meaning for 'dest' and 'src'? > Answering to myself... yes. in that commit. > Better commit log would have made the code review so much easier. > > > skl_call_thunk_template, tmpl_size); > > > > memcpy(*pprog, insn_buff, tmpl_size); > > diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c > > index 09f7dc9d4d65..f2e8769f5eee 100644 > > --- a/arch/x86/net/bpf_jit_comp.c > > +++ b/arch/x86/net/bpf_jit_comp.c > > @@ -481,7 +481,7 @@ static int emit_rsb_call(u8 **pprog, void *func, void *ip) > > { > > void *adjusted_ip; > > OPTIMIZER_HIDE_VAR(func); > > - adjusted_ip = ip + x86_call_depth_emit_accounting(pprog, func); > > + adjusted_ip = ip + x86_call_depth_emit_accounting(pprog, func, ip); > > Now I see why you added extra var in the previous patch. > Should have mentioned it in the commit log. > > > return emit_patch(pprog, func, adjusted_ip, 0xE8); > > } > > > > @@ -1973,20 +1973,17 @@ st: if (is_imm8(insn->off)) > > > > /* call */ > > case BPF_JMP | BPF_CALL: { > > - int offs; > > + u8 *ip = image + addrs[i - 1]; > > > > func = (u8 *) __bpf_call_base + imm32; > > if (tail_call_reachable) { > > RESTORE_TAIL_CALL_CNT(bpf_prog->aux->stack_depth); > > - if (!imm32) > > - return -EINVAL; > > - offs = 7 + x86_call_depth_emit_accounting(&prog, func); > > - } else { > > - if (!imm32) > > - return -EINVAL; > > - offs = x86_call_depth_emit_accounting(&prog, func); > > + ip += 7; > > } > > - if (emit_call(&prog, func, image + addrs[i - 1] + offs)) > > + if (!imm32) > > + return -EINVAL; > > + ip += x86_call_depth_emit_accounting(&prog, func, ip); > > + if (emit_call(&prog, func, ip)) > > return -EINVAL; > > break; > > } > > @@ -2836,7 +2833,7 @@ static int __arch_prepare_bpf_trampoline(struct bpf_tramp_image *im, void *rw_im > > * Direct-call fentry stub, as such it needs accounting for the > > * __fentry__ call. > > */ > > - x86_call_depth_emit_accounting(&prog, NULL); > > + x86_call_depth_emit_accounting(&prog, NULL, image); > > Overall it all makes sense. > Pls respin with more precise commit logs.
On Sat, Mar 30, 2024 at 2:01 AM Uros Bizjak <ubizjak@gmail.com> wrote: > > On Fri, Mar 29, 2024 at 10:53 PM Alexei Starovoitov > <alexei.starovoitov@gmail.com> wrote: > > > > On Fri, Mar 29, 2024 at 2:49 AM Uros Bizjak <ubizjak@gmail.com> wrote: > > > > > > From: Joan Bruguera Micó <joanbrugueram@gmail.com> > > > > > > The recently introduced support for %rip-relative relocations in the > > > call thunk template assumes that the code is being patched in-place, > > > so the destination of the relocation matches the address of the code. > > > This is not true for the call depth accounting emitted by the BPF JIT, > > > so the calculated address is wrong and usually causes a page fault. > > > > Could you share the link to what this 'rip-relative' relocation is ? > > Please see the "RIP relative addressing" section in [1]. > > [1] https://compas.cs.stonybrook.edu/~nhonarmand/courses/sp17/cse506/ref/assembly.html > > In our case: > > The callthunks patching creates a call thunk template in the .rodata > section (please see arch/x86/kernel/callthunks.c) that is later > copied to the .text section at the correct place. The template uses > X86_call_depth in the pcpu_hot structure. Previously, the template > used absolute location for X86_call_depth and the linker resolved the > address in the template to this absolute location. There is no issue > when this template is copied to the various places in the .text > section. > > When we want to use PC relative relocations (to reduce the code size), > then the linker calculates the address of the variable in the template > according to the PC in the .rodata section. If we want to copy the > template to its final location, then the address of X86_call_depth, > relative to the PC, has to be adjusted, as explained in > arch/x86/kernel/alternative.c, in the comment above apply_reloc_n > macro. I didn't mean to ask for info about the definition of rip-relative, but how it's used in this case and what you've been trying to achieve with commit 17bce3b2ae2d that broke call depth accounting. And the whole sequence of events that caused this breakage. Something like: commit 59bec00ace28 ("x86/percpu: Introduce %rip-relative addressing to PER_CPU_VAR()") made PER_CPU_VAR() to use rip-relative addressing, hence INCREMENT_CALL_DEPTH macro and skl_call_thunk_template got rip-relative asm code inside of it. Hence x86_call_depth_emit_accounting() was changed in commit 17bce3b2ae2d ("x86/callthunks: Handle %rip-relative relocations in call thunk template") to use apply_relocation(), but it was mistakenly made to use *pprog as dest ip, so jit-ed bpf progs on kernels with call depth tracking got broken. Such details should be in the commit log.
On Mon, Apr 1, 2024 at 8:03 PM Alexei Starovoitov <alexei.starovoitov@gmail.com> wrote: > > On Sat, Mar 30, 2024 at 2:01 AM Uros Bizjak <ubizjak@gmail.com> wrote: > > > > On Fri, Mar 29, 2024 at 10:53 PM Alexei Starovoitov > > <alexei.starovoitov@gmail.com> wrote: > > > > > > On Fri, Mar 29, 2024 at 2:49 AM Uros Bizjak <ubizjak@gmail.com> wrote: > > > > > > > > From: Joan Bruguera Micó <joanbrugueram@gmail.com> > > > > > > > > The recently introduced support for %rip-relative relocations in the > > > > call thunk template assumes that the code is being patched in-place, > > > > so the destination of the relocation matches the address of the code. > > > > This is not true for the call depth accounting emitted by the BPF JIT, > > > > so the calculated address is wrong and usually causes a page fault. > > > > > > Could you share the link to what this 'rip-relative' relocation is ? > > > > Please see the "RIP relative addressing" section in [1]. > > > > [1] https://compas.cs.stonybrook.edu/~nhonarmand/courses/sp17/cse506/ref/assembly.html > > > > In our case: > > > > The callthunks patching creates a call thunk template in the .rodata > > section (please see arch/x86/kernel/callthunks.c) that is later > > copied to the .text section at the correct place. The template uses > > X86_call_depth in the pcpu_hot structure. Previously, the template > > used absolute location for X86_call_depth and the linker resolved the > > address in the template to this absolute location. There is no issue > > when this template is copied to the various places in the .text > > section. > > > > When we want to use PC relative relocations (to reduce the code size), > > then the linker calculates the address of the variable in the template > > according to the PC in the .rodata section. If we want to copy the > > template to its final location, then the address of X86_call_depth, > > relative to the PC, has to be adjusted, as explained in > > arch/x86/kernel/alternative.c, in the comment above apply_reloc_n > > macro. > > I didn't mean to ask for info about the definition of rip-relative, > but how it's used in this case and what you've been trying > to achieve with commit 17bce3b2ae2d that broke call depth accounting. > And the whole sequence of events that caused this breakage. > Something like: > commit 59bec00ace28 ("x86/percpu: Introduce %rip-relative addressing > to PER_CPU_VAR()") > made PER_CPU_VAR() to use rip-relative addressing, > hence INCREMENT_CALL_DEPTH macro and skl_call_thunk_template > got rip-relative asm code inside of it. > Hence x86_call_depth_emit_accounting() was changed > in commit 17bce3b2ae2d ("x86/callthunks: Handle %rip-relative > relocations in call thunk template") to use apply_relocation(), > but it was mistakenly made to use *pprog as dest ip, > so jit-ed bpf progs on kernels with call depth tracking got broken. > Such details should be in the commit log. Oh, I was not sure that all those x86 specific details should be in the commit log, since x86 maintainers already acked the patch. Sure, I'll add your description of the fix to the patch commit message, it really describes the problem in a way, understandable also to non-x86 people. Thanks, Uros.
diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index fcd20c6dc7f9..67b68d0d17d1 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -117,7 +117,7 @@ extern void callthunks_patch_builtin_calls(void); extern void callthunks_patch_module_calls(struct callthunk_sites *sites, struct module *mod); extern void *callthunks_translate_call_dest(void *dest); -extern int x86_call_depth_emit_accounting(u8 **pprog, void *func); +extern int x86_call_depth_emit_accounting(u8 **pprog, void *func, void *ip); #else static __always_inline void callthunks_patch_builtin_calls(void) {} static __always_inline void @@ -128,7 +128,7 @@ static __always_inline void *callthunks_translate_call_dest(void *dest) return dest; } static __always_inline int x86_call_depth_emit_accounting(u8 **pprog, - void *func) + void *func, void *ip) { return 0; } diff --git a/arch/x86/kernel/callthunks.c b/arch/x86/kernel/callthunks.c index 30335182b6b0..e92ff0c11db8 100644 --- a/arch/x86/kernel/callthunks.c +++ b/arch/x86/kernel/callthunks.c @@ -314,7 +314,7 @@ static bool is_callthunk(void *addr) return !bcmp(pad, insn_buff, tmpl_size); } -int x86_call_depth_emit_accounting(u8 **pprog, void *func) +int x86_call_depth_emit_accounting(u8 **pprog, void *func, void *ip) { unsigned int tmpl_size = SKL_TMPL_SIZE; u8 insn_buff[MAX_PATCH_LEN]; @@ -327,7 +327,7 @@ int x86_call_depth_emit_accounting(u8 **pprog, void *func) return 0; memcpy(insn_buff, skl_call_thunk_template, tmpl_size); - apply_relocation(insn_buff, tmpl_size, *pprog, + apply_relocation(insn_buff, tmpl_size, ip, skl_call_thunk_template, tmpl_size); memcpy(*pprog, insn_buff, tmpl_size); diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index 09f7dc9d4d65..f2e8769f5eee 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -481,7 +481,7 @@ static int emit_rsb_call(u8 **pprog, void *func, void *ip) { void *adjusted_ip; OPTIMIZER_HIDE_VAR(func); - adjusted_ip = ip + x86_call_depth_emit_accounting(pprog, func); + adjusted_ip = ip + x86_call_depth_emit_accounting(pprog, func, ip); return emit_patch(pprog, func, adjusted_ip, 0xE8); } @@ -1973,20 +1973,17 @@ st: if (is_imm8(insn->off)) /* call */ case BPF_JMP | BPF_CALL: { - int offs; + u8 *ip = image + addrs[i - 1]; func = (u8 *) __bpf_call_base + imm32; if (tail_call_reachable) { RESTORE_TAIL_CALL_CNT(bpf_prog->aux->stack_depth); - if (!imm32) - return -EINVAL; - offs = 7 + x86_call_depth_emit_accounting(&prog, func); - } else { - if (!imm32) - return -EINVAL; - offs = x86_call_depth_emit_accounting(&prog, func); + ip += 7; } - if (emit_call(&prog, func, image + addrs[i - 1] + offs)) + if (!imm32) + return -EINVAL; + ip += x86_call_depth_emit_accounting(&prog, func, ip); + if (emit_call(&prog, func, ip)) return -EINVAL; break; } @@ -2836,7 +2833,7 @@ static int __arch_prepare_bpf_trampoline(struct bpf_tramp_image *im, void *rw_im * Direct-call fentry stub, as such it needs accounting for the * __fentry__ call. */ - x86_call_depth_emit_accounting(&prog, NULL); + x86_call_depth_emit_accounting(&prog, NULL, image); } EMIT1(0x55); /* push rbp */ EMIT3(0x48, 0x89, 0xE5); /* mov rbp, rsp */