Supporting New Memory Barrier Types in BPF
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPF
To that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------
Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins like
__sync_val_compare_and_swap()
We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
II. Implicit barriers can get confusing
---------------------------------------
We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used __sync_fetch_and_add() can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:
Case (a): return value is used
SEC("...")
int64_t foo;
int64_t func(...) {
return __sync_fetch_and_add(&foo, 1);
}For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 = atomic_fetch_add((u64 *)(r1 + 0x0), r0)
opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW
imm (0x00000001): BPF_ADD | BPF_FETCH
Case (b): return value is ignored
SEC("...")
int64_t foo;
int64_t func(...) {
__sync_fetch_and_add(&foo, 1);return foo;
}
For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 + 0x0) += r1
opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW
imm (0x00000000): BPF_ADD
LLVM decided to drop BPF_FETCH, since the return value of __sync_fetch_and_add() is being ignored [2]. Now, if we take a look at emit_lse_atomic() in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):
case BPF_ADD:
emit(A64_STADD(isdw, reg, src), ctx);
break;
<...>
case BPF_ADD | BPF_FETCH:
emit(A64_LDADDAL(isdw, src, reg, src), ctx);
break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"
This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
type __atomic_fetch_add (type *ptr, type val, int memorder)
This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
III. Next steps
---------------
Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
* teach LLVM and GCC to generate the new/extended instructions
* teach the BPF verifier to understand them
* teach the BPF JIT compiler to compile them
* update BPF memory model and tooling
* update IETF specification
Additionally, for the issue described in the previous section, we need to:
* check if GCC has the same behavior
* at least clearly document the implied effects on BPF memory ordering of current __sync_*fetch*() built-ins (especially for architectures like ARM), as described
* fully support the new __atomic_*fetch*() built-ins for BPF to replace the
__sync_*fetch*() ones
Any suggestions or corrections would be most welcome!
Thanks,
Peilin Ye
[1] Instruction-Level BPF Memory Model
https://docs.google.com/document/d/1TaSEfWfLnRUi5KqkavUQyL2tThJXYWHS15qcbxIsFb0/edit?usp=sharing
[2] For more information, see LLVM commit 286daafd6512 ("[BPF] support atomic instructions"). Search for "LLVM will check the return value" in the commit message.
[3] Arm Architecture Reference Manual for A-profile architecture (ARM DDI
0487K.a, ID032224), C6.2.149, page 2006
[4] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fsync-Builtins.html
6.58 Legacy __sync Built-in Functions for Atomic Memory Access "In most cases, these built-in functions are considered a full barrier."
[5] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html
6.59 Built-in Functions for Memory Model Aware Atomic Operations
Quoted 25 lines · Peilin Ye
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins like
Quoted 7 lines
__sync_val_compare_and_swap()We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);
we would need inline asm equivalent too. Similar to kernel smp_load_acquire() macro.
Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.
Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...")
int64_tfoo;int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 =
atomic_fetch_add((u64 *)(r1 + 0x0), r0)opcode (
0xdb):BPF_STX|BPF_ATOMIC|BPF_DW
imm (0x00000001):BPF_ADD|BPF_FETCHCase (b): return value is ignored
SEC("...")
int64_tfoo;int64_t func(...) { __sync_fetch_and_add(&foo, 1);return foo;}
For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 +
0x0) += r1opcode (
0xdb):BPF_STX|BPF_ATOMIC|BPF_DW
imm (0x00000000):BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convert sycn_fetch_add into sync_add when fetch part is unused?
Quoted 7 lines
III. Next steps
---------------Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding MEM_ACQ_REL feels like a natural fit. Better name?
For barriers we would need a new insn. Not sure which class would fit the best. Maybe BPF_LD ?
Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel code smp_mb(), smp_rmb(), etc.
I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Quoted 38 lines
* teach LLVM and GCC to generate the new/extended instructions
* teach the BPF verifier to understand them
* teach the BPF JIT compiler to compile them
* update BPF memory model and tooling
* update IETF specificationAdditionally, for the issue described in the previous section, we need to:
* check if GCC has the same behavior
* at least clearly document the implied effects on BPF memory ordering of current __sync_*fetch*() built-ins (especially for architectures like ARM), as described
* fully support the new __atomic_*fetch*() built-ins for BPF to replace the__sync_*fetch*() onesAny suggestions or corrections would be most welcome!
Thanks,
Peilin Ye[1] Instruction-Level BPF Memory Model
https://docs.google.com/document/d/1TaSEfWfLnRUi5KqkavUQyL2tThJXYWHS15qcbxIsFb0/edit?usp=sharing[2] For more information, see LLVM commit
286daafd6512("[BPF] support atomic instructions"). Search for "LLVM will check the return value" in the commit message.[3] Arm Architecture Reference Manual for A-profile architecture (ARM DDI
0487K.a, ID032224), C6.2.149, page 2006[4] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fsync-Builtins.html
6.58 Legacy__syncBuilt-in Functions for Atomic Memory Access "In most cases, these built-in functions are considered a full barrier."[5] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html
6.59 Built-in Functions for Memory Model Aware Atomic Operations
Quoted 43 lines · Alexei Starovoitov
Quoted 25 lines · Peilin Ye
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins likeQuoted 7 lines
__sync_val_compare_and_swap()We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
Quoted 6 lines
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.
When reading BPF instructions back into a compiler backend, would it make sense to convert an acquire-load instruction back to __atomic_load_n(... memorder=__ATOMIC_ACQUIRE)? Or the internal representation thereof?
Thanx, Paul
Quoted 147 lines
Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...")
int64_tfoo;int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 =
atomic_fetch_add((u64 *)(r1 + 0x0), r0)opcode (
0xdb):BPF_STX|BPF_ATOMIC|BPF_DW
imm (0x00000001):BPF_ADD|BPF_FETCHCase (b): return value is ignored
SEC("...")
int64_tfoo;int64_t func(...) { __sync_fetch_and_add(&foo, 1);return foo;}
For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 +
0x0) += r1opcode (
0xdb):BPF_STX|BPF_ATOMIC|BPF_DW
imm (0x00000000):BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?Quoted 7 lines
III. Next steps
---------------Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?For barriers we would need a new insn. Not sure which class would fit the best. Maybe
BPF_LD?Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel codesmp_mb(),smp_rmb(), etc.I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Quoted 38 lines
* teach LLVM and GCC to generate the new/extended instructions
* teach the BPF verifier to understand them
* teach the BPF JIT compiler to compile them
* update BPF memory model and tooling
* update IETF specificationAdditionally, for the issue described in the previous section, we need to:
* check if GCC has the same behavior
* at least clearly document the implied effects on BPF memory ordering of current __sync_*fetch*() built-ins (especially for architectures like ARM), as described
* fully support the new __atomic_*fetch*() built-ins for BPF to replace the__sync_*fetch*() onesAny suggestions or corrections would be most welcome!
Thanks,
Peilin Ye[1] Instruction-Level BPF Memory Model
https://docs.google.com/document/d/1TaSEfWfLnRUi5KqkavUQyL2tThJXYWHS15qcbxIsFb0/edit?usp=sharing[2] For more information, see LLVM commit
286daafd6512("[BPF] support atomic instructions"). Search for "LLVM will check the return value" in the commit message.[3] Arm Architecture Reference Manual for A-profile architecture (ARM DDI
0487K.a, ID032224), C6.2.149, page 2006[4] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fsync-Builtins.html
6.58 Legacy__syncBuilt-in Functions for Atomic Memory Access "In most cases, these built-in functions are considered a full barrier."[5] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html
6.59 Built-in Functions for Memory Model Aware Atomic Operations
Quoted 9 lines · Paul E. McKenney
Quoted 8 lines
Quoted 1 line
instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.
Quoted 4 lines
When reading BPF instructions back into a compiler backend, would it make sense to convert an acquire-load instruction back to __atomic_load_n(... memorder=__ATOMIC_ACQUIRE)? Or the internal representation thereof?
Internal representation?
We need to pick asm mnemonics for ld_acq insn, but pushing C-like asm to emit __atomic_load_n() in disasm is imo too much. Currently LDX insn will be disasm-ed to r1 = *(u32 *)r2; For load acquire insn we can emit r1 = smp_load_acquire_u32(r2); or r1 = load_acquire((u32 *)r2);
Yonghong Song
reviewerQuoted 125 lines · Alexei Starovoitov
Quoted 24 lines · Peilin Ye
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins likeQuoted 7 lines
__sync_val_compare_and_swap()We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...") int64_t foo; int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 = atomic_fetch_add((u64 *)(r1 + 0x0), r0) opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000001): BPF_ADD | BPF_FETCHCase (b): return value is ignored
SEC("...") int64_t foo; int64_t func(...) { __sync_fetch_and_add(&foo, 1); return foo; }For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 + 0x0) += r1 opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000000): BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?
This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744
Specifically the following commit message:
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.
======
Basically, if no return value, __sync_fetch_and_add() will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we have
#define lock_xadd(ptr, val) ((void)__sync_fetch_and_add(ptr, val))
Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
Quoted 62 lines
Quoted 8 lines
III. Next steps
---------------Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?For barriers we would need a new insn. Not sure which class would fit the best. Maybe
BPF_LD?Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel codesmp_mb(),smp_rmb(), etc.I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Quoted 38 lines
* teach LLVM and GCC to generate the new/extended instructions
* teach the BPF verifier to understand them
* teach the BPF JIT compiler to compile them
* update BPF memory model and tooling
* update IETF specificationAdditionally, for the issue described in the previous section, we need to:
* check if GCC has the same behavior
* at least clearly document the implied effects on BPF memory ordering of current __sync_*fetch*() built-ins (especially for architectures like ARM), as described
* fully support the new __atomic_*fetch*() built-ins for BPF to replace the__sync_*fetch*() onesAny suggestions or corrections would be most welcome!
Thanks,
Peilin Ye[1] Instruction-Level BPF Memory Model
https://docs.google.com/document/d/1TaSEfWfLnRUi5KqkavUQyL2tThJXYWHS15qcbxIsFb0/edit?usp=sharing[2] For more information, see LLVM commit
286daafd6512("[BPF] support atomic instructions"). Search for "LLVM will check the return value" in the commit message.[3] Arm Architecture Reference Manual for A-profile architecture (ARM DDI
0487K.a, ID032224), C6.2.149, page 2006[4] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fsync-Builtins.html
6.58 Legacy__syncBuilt-in Functions for Atomic Memory Access "In most cases, these built-in functions are considered a full barrier."[5] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html
6.59 Built-in Functions for Memory Model Aware Atomic Operations
Quoted 12 lines · Yonghong Song
Quoted 6 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
Quoted 7 lines
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.
So it's a bpf backend bug. Great. That's fixable. Would have been much harder if this transformation was performed by the middle end.
Quoted 9 lines
======
Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we have
#definelock_xadd(ptr, val)((void)__sync_fetch_and_add(ptr, val))
Quoted 3 lines
Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
Right. We did it for backward compat. Older llvm was completely wrong to generate xadd for __sync_fetch_and_add. That was my hack from 10 years ago when xadd was all we had. So we fixed that old llvm bug, but introduced another with all good intentions.
Since proper atomic insns were introduced 3 years ago we should remove this backward compat feature/bug from llvm. The only breakage is for kernels older than 5.12.
I think that's an acceptable risk.
Yonghong Song
reviewerQuoted 34 lines · Alexei Starovoitov
Quoted 10 lines · Yonghong Song
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
Quoted 7 lines
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.So it's a bpf backend bug. Great. That's fixable. Would have been much harder if this transformation was performed by the middle end.
Quoted 7 lines
======
Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we haveQuoted 5 lines
#define lock_xadd(ptr, val) ((void)__sync_fetch_and_add(ptr, val))Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
Quoted 9 lines
Right. We did it for backward compat. Older llvm was completely wrong to generate xadd for
__sync_fetch_and_add. That was my hack from 10 years ago when xadd was all we had. So we fixed that old llvm bug, but introduced another with all good intentions.
Since proper atomic insns were introduced 3 years ago we should remove this backward compat feature/bug from llvm. The only breakage is for kernels older than 5.12.
I think that's an acceptable risk.
Sounds good, I will remove the backward compat part in llvm20.
Hi Alexei, Yonghong,
Quoted 46 lines · Yonghong Song
Quoted 36 lines
Quoted 11 lines
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
Quoted 7 lines
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.So it's a bpf backend bug. Great. That's fixable. Would have been much harder if this transformation was performed by the middle end.
Quoted 7 lines
======
Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we haveQuoted 5 lines
#define lock_xadd(ptr, val) ((void)__sync_fetch_and_add(ptr, val))Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
Quoted 9 lines
Right. We did it for backward compat. Older llvm was completely wrong to generate xadd for
__sync_fetch_and_add. That was my hack from 10 years ago when xadd was all we had. So we fixed that old llvm bug, but introduced another with all good intentions.
Since proper atomic insns were introduced 3 years ago we should remove this backward compat feature/bug from llvm. The only breakage is for kernels older than 5.12.
I think that's an acceptable risk.
Quoted 1 line
Sounds good, I will remove the backward compat part in llvm20.
Thanks for confirming! Would you mind if I fix it myself? It may affect some of the BPF code that we will be running on ARM, so we would like to get it fixed sooner. Also, I would love to gain some experience in LLVM development!
Thanks,
Peilin Ye
Yonghong Song
reviewerQuoted 45 lines · Peilin Ye
Hi Alexei, Yonghong,
Quoted 42 lines · Yonghong Song
Quoted 42 lines
Quoted 10 lines
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
Quoted 7 lines
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.So it's a bpf backend bug. Great. That's fixable. Would have been much harder if this transformation was performed by the middle end.
Quoted 7 lines
======
Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we haveQuoted 5 lines
#define lock_xadd(ptr, val) ((void)__sync_fetch_and_add(ptr, val))Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
Right. We did it for backward compat. Older llvm was completely wrong to generate xadd for
__sync_fetch_and_add. That was my hack from 10 years ago when xadd was all we had. So we fixed that old llvm bug, but introduced another with all good intentions.
Since proper atomic insns were introduced 3 years ago we should remove this backward compat feature/bug from llvm. The only breakage is for kernels older than 5.12.
I think that's an acceptable risk.
Quoted 5 lines
Quoted 1 line
Sounds good, I will remove the backward compat part in llvm20.
Thanks for confirming! Would you mind if I fix it myself? It may affect some of the BPF code that we will be running on ARM, so we would like to get it fixed sooner. Also, I would love to gain some experience in LLVM development!
Peilin, when I saw your email, I have almost done with the change. The below is the llvm patch:
https://github.com/llvm/llvm-project/pull/101428
Please help take a look. You are certainly welcome to do llvm related work. Just respond earlier to mention you intend to do a particular llvm patch and we are happy for you to contribute and will help when you have any questions.
Quoted 4 lines
Thanks,
Peilin Ye
Hi Yonghong,
Quoted 9 lines · Yonghong Song
Quoted 4 lines · Peilin Ye
Thanks for confirming! Would you mind if I fix it myself? It may affect some of the BPF code that we will be running on ARM, so we would like to get it fixed sooner. Also, I would love to gain some experience in LLVM development!
Peilin, when I saw your email, I have almost done with the change. The below is the llvm patch:
https://github.com/llvm/llvm-project/pull/101428
Wow, that is really fast! Thanks for the quick fix.
Quoted 4 lines
Please help take a look. You are certainly welcome to do llvm related work. Just respond earlier to mention you intend to do a particular llvm patch and we are happy for you to contribute and will help when you have any questions.
Sure! I'll look into that PR.
Thanks,
Peilin Ye
Quoted 154 lines · Alexei Starovoitov
Quoted 125 lines · Peilin Ye
Quoted 24 lines
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins likeQuoted 7 lines
__sync_val_compare_and_swap()We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...") int64_t foo; int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 = atomic_fetch_add((u64 *)(r1 + 0x0), r0) opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000001): BPF_ADD | BPF_FETCHCase (b): return value is ignored
SEC("...") int64_t foo; int64_t func(...) { __sync_fetch_and_add(&foo, 1); return foo; }For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 + 0x0) += r1 opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000000): BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.
======Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we have
#definelock_xadd(ptr, val)((void)__sync_fetch_and_add(ptr, val))
Quoted 3 lines
Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
GCC behaves similarly.
For program A:
long foo;
long func () {
return __sync_fetch_and_add(&foo, 1);
}bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>:
0: 18 00 00 00 00 00 00 00 r0=0 ll
8: 00 00 00 00 00 00 00 00
10: b7 01 00 00 01 00 00 00 r1=1
18: db 10 00 00 01 00 00 00 r1=atomic_fetch_add((u64*)(r0+0),r1)
20: bf 10 00 00 00 00 00 00 r0=r1
28: 95 00 00 00 00 00 00 00 exitAnd for program B:
long foo;
long func () {
__sync_fetch_and_add(&foo, 1);
return foo;
}bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>:
0: 18 00 00 00 00 00 00 00 r0=0 ll
8: 00 00 00 00 00 00 00 00
10: b7 01 00 00 01 00 00 00 r1=1
18: db 10 00 00 00 00 00 00 lock *(u64*)(r0+0)+=r1
20: 79 00 00 00 00 00 00 00 r0=*(u64*)(r0+0)
28: 95 00 00 00 00 00 00 00 exitInternally:
- When compiling the program A GCC decides to emit an `atomic_fetch_addDI' insn, documented as:
'atomic_fetch_addMODE', 'atomic_fetch_subMODE'
'atomic_fetch_orMODE', 'atomic_fetch_andMODE'
'atomic_fetch_xorMODE', 'atomic_fetch_nandMODE'
These patterns emit code for an atomic operation on memory with memory model semantics, and return the original value. Operand 0 is an output operand which contains the value of the memory location before the operation was performed. Operand 1 is the memory on which the atomic operation is performed. Operand 2 is the second operand to the binary operator. Operand 3 is the memory model to be used by the operation.
The BPF backend defines atomic_fetch_add for DI modes (long) to expand to this BPF instruction:
%w0 = atomic_fetch_add((<smop> *)%1, %w0)
- When compiling the program B GCC decides to emit an `atomic_addDI' insn, documented as:
'atomic_addMODE', 'atomic_subMODE'
'atomic_orMODE', 'atomic_andMODE'
'atomic_xorMODE', 'atomic_nandMODE'
These patterns emit code for an atomic operation on memory with memory model semantics. Operand 0 is the memory on which the atomic operation is performed. Operand 1 is the second operand to the binary operator. Operand 2 is the memory model to be used by the operation.
The BPF backend defines atomic_fetch_add for DI modes (long) to expand to this BPF instruction:
lock *(<smop> *)%w0 += %w1
This is done for all targets. In x86-64, for example, case A compiles to:
0000000000000000 <func>:
0: b8 01 00 00 00 mov $0x1,%eax
5: f0 48 0f c1 05 00 00 lock xadd %rax,0x0(%rip) # e <func+0xe>
c: 00 00
e: c3 retq And case B compiles to:
0000000000000000 <func>:
0: f0 48 83 05 00 00 00 lock addq $0x1,0x0(%rip) # 9 <func+0x9>
7: 00 01
9: 48 8b 05 00 00 00 00 mov 0x0(%rip),%rax # 10 <func+0x10>
10: c3 retq Why wouldn't the compiler be allowed to optimize from atomic_fetch_add to atomic_add in this case?
Quoted 63 lines
Quoted 63 lines
Quoted 9 lines
III. Next steps
---------------Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?For barriers we would need a new insn. Not sure which class would fit the best. Maybe
BPF_LD?Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel codesmp_mb(),smp_rmb(), etc.I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Quoted 38 lines
* teach LLVM and GCC to generate the new/extended instructions
* teach the BPF verifier to understand them
* teach the BPF JIT compiler to compile them
* update BPF memory model and tooling
* update IETF specificationAdditionally, for the issue described in the previous section, we need to:
* check if GCC has the same behavior
* at least clearly document the implied effects on BPF memory ordering of current __sync_*fetch*() built-ins (especially for architectures like ARM), as described
* fully support the new __atomic_*fetch*() built-ins for BPF to replace the__sync_*fetch*() onesAny suggestions or corrections would be most welcome!
Thanks,
Peilin Ye[1] Instruction-Level BPF Memory Model
https://docs.google.com/document/d/1TaSEfWfLnRUi5KqkavUQyL2tThJXYWHS15qcbxIsFb0/edit?usp=sharing[2] For more information, see LLVM commit
286daafd6512("[BPF] support atomic instructions"). Search for "LLVM will check the return value" in the commit message.[3] Arm Architecture Reference Manual for A-profile architecture (ARM DDI
0487K.a, ID032224), C6.2.149, page 2006[4] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fsync-Builtins.html
6.58 Legacy__syncBuilt-in Functions for Atomic Memory Access "In most cases, these built-in functions are considered a full barrier."[5] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html
6.59 Built-in Functions for Memory Model Aware Atomic Operations
Quoted 248 lines
Quoted 154 lines · Alexei Starovoitov
Quoted 125 lines · Peilin Ye
Quoted 24 lines
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins likeQuoted 7 lines
__sync_val_compare_and_swap()We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...") int64_t foo; int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 = atomic_fetch_add((u64 *)(r1 + 0x0), r0) opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000001): BPF_ADD | BPF_FETCHCase (b): return value is ignored
SEC("...") int64_t foo; int64_t func(...) { __sync_fetch_and_add(&foo, 1); return foo; }For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 + 0x0) += r1 opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000000): BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.
======Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we have
#definelock_xadd(ptr, val)((void)__sync_fetch_and_add(ptr, val))Quoted 3 lines
Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
GCC behaves similarly.
For program A:
long foo;
long func () { return __sync_fetch_and_add(&foo, 1); }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>:
0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 01 00 00 00 r1=atomic_fetch_add((u64*)(r0+0),r1) 20: bf 10 00 00 00 00 00 00 r0=r1 28: 95 00 00 00 00 00 00 00 exitAnd for program B:
long foo;
long func () { __sync_fetch_and_add(&foo, 1); return foo; }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>:
0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 00 00 00 00 lock *(u64*)(r0+0)+=r1 20: 79 00 00 00 00 00 00 00 r0=*(u64*)(r0+0) 28: 95 00 00 00 00 00 00 00 exitInternally:
- When compiling the program A GCC decides to emit an `atomic_fetch_addDI' insn, documented as:
'
atomic_fetch_addMODE', 'atomic_fetch_subMODE'
'atomic_fetch_orMODE', 'atomic_fetch_andMODE'
'atomic_fetch_xorMODE', 'atomic_fetch_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics, and return the original value. Operand 0 is an output operand which contains the value of the memory location before the operation was performed. Operand 1 is the memory on which the atomic operation is performed. Operand 2 is the second operand to the binary operator. Operand 3 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:%w0 = atomic_fetch_add((<smop> *)%1, %w0)- When compiling the program B GCC decides to emit an `atomic_addDI' insn, documented as:
'
atomic_addMODE', 'atomic_subMODE'
'atomic_orMODE', 'atomic_andMODE'
'atomic_xorMODE', 'atomic_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics. Operand 0 is the memory on which the atomic operation is performed. Operand 1 is the second operand to the binary operator. Operand 2 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:lock *(<smop> *)%w0 += %w1This is done for all targets. In x86-64, for example, case A compiles to:
0000000000000000 <func>:
0: b8 01 00 00 00 mov $0x1,%eax 5: f0 48 0f c1 05 00 00 lock xadd %rax,0x0(%rip) # e <func+0xe> c: 00 00 e: c3 retqAnd case B compiles to:
Quoted 8 lines
0000000000000000 <func>:
0: f0 48 83 05 00 00 00 lock addq $0x1,0x0(%rip) # 9 <func+0x9> 7: 00 01 9: 48 8b 05 00 00 00 00 mov 0x0(%rip),%rax # 10 <func+0x10> 10: c3 retqWhy wouldn't the compiler be allowed to optimize from
atomic_fetch_addtoatomic_addin this case?
Ok I see. The generic compiler optimization is ok. It is the backend that is buggy because it emits BPF instruction sequences with different memory ordering semantics for atomic_OP and atomic_fetch_OP.
The only difference between fetching and non-fetching builtins is that in one case the original value is returned, in the other the new value. Other than that they should be equivalent.
For ARM64, GCC generates for case A:
0000000000000000 <func>:
0: 90000001 adrp x1, 0 <func>
4: d2800020 mov x0, #0x1 // #1
8: 91000021 add x1, x1, #0x0
c: f8e00020 ldaddal x0, x0, [x1]
10: d65f03c0 retAnd this for case B:
0000000000000000 <func>:
0: 90000000 adrp x0, 0 <func>
4: d2800022 mov x2, #0x1 // #1
8: 91000001 add x1, x0, #0x0
c: f8e20021 ldaddal x2, x1, [x1]
10: f9400000 ldr x0, [x0]
14: d65f03c0 reti.e. GCC emits LDADDAL for both atomic_add and atomic_fetch_add internal insns. Like in x86-64, both sequences have same memory ordering semantics.
Allright we are changing GCC to always emit fetch versions of sequences for all the supported atomic operations: add, and, or, xor. After the change the `lock' versions of the instructions will not be generated by the compiler at all out of inline asm.
Will send a headsup when done.
Salud!
Quoted 62 lines
Quoted 62 lines
Quoted 62 lines
Quoted 8 lines
III. Next steps
---------------Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?For barriers we would need a new insn. Not sure which class would fit the best. Maybe
BPF_LD?Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel codesmp_mb(),smp_rmb(), etc.I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Quoted 38 lines
* teach LLVM and GCC to generate the new/extended instructions
* teach the BPF verifier to understand them
* teach the BPF JIT compiler to compile them
* update BPF memory model and tooling
* update IETF specificationAdditionally, for the issue described in the previous section, we need to:
* check if GCC has the same behavior
* at least clearly document the implied effects on BPF memory ordering of current __sync_*fetch*() built-ins (especially for architectures like ARM), as described
* fully support the new __atomic_*fetch*() built-ins for BPF to replace the__sync_*fetch*() onesAny suggestions or corrections would be most welcome!
Thanks,
Peilin Ye[1] Instruction-Level BPF Memory Model
https://docs.google.com/document/d/1TaSEfWfLnRUi5KqkavUQyL2tThJXYWHS15qcbxIsFb0/edit?usp=sharing[2] For more information, see LLVM commit
286daafd6512("[BPF] support atomic instructions"). Search for "LLVM will check the return value" in the commit message.[3] Arm Architecture Reference Manual for A-profile architecture (ARM DDI
0487K.a, ID032224), C6.2.149, page 2006[4] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fsync-Builtins.html
6.58 Legacy__syncBuilt-in Functions for Atomic Memory Access "In most cases, these built-in functions are considered a full barrier."[5] https://gcc.gnu.org/onlinedocs/gcc/_005f_005fatomic-Builtins.html
6.59 Built-in Functions for Memory Model Aware Atomic Operations
Yonghong Song
reviewerQuoted 284 lines · Jose E. Marchesi
Quoted 245 lines
Quoted 152 lines · Alexei Starovoitov
Quoted 125 lines · Peilin Ye
Quoted 26 lines
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins like__sync_val_compare_and_swap()Quoted 5 lines
We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...") int64_t foo; int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 = atomic_fetch_add((u64 *)(r1 + 0x0), r0) opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000001): BPF_ADD | BPF_FETCHCase (b): return value is ignored
SEC("...") int64_t foo; int64_t func(...) { __sync_fetch_and_add(&foo, 1); return foo; }For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 + 0x0) += r1 opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000000): BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.
======Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we have
#definelock_xadd(ptr, val)((void)__sync_fetch_and_add(ptr, val))Quoted 3 lines
Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
GCC behaves similarly.
For program A:
long foo; long func () { return __sync_fetch_and_add(&foo, 1); }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>: 0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 01 00 00 00 r1=atomic_fetch_add((u64*)(r0+0),r1) 20: bf 10 00 00 00 00 00 00 r0=r1 28: 95 00 00 00 00 00 00 00 exitAnd for program B:
long foo; long func () { __sync_fetch_and_add(&foo, 1); return foo; }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>: 0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 00 00 00 00 lock *(u64*)(r0+0)+=r1 20: 79 00 00 00 00 00 00 00 r0=*(u64*)(r0+0) 28: 95 00 00 00 00 00 00 00 exitInternally:
- When compiling the program A GCC decides to emit an `atomic_fetch_addDI' insn, documented as:
'
atomic_fetch_addMODE', 'atomic_fetch_subMODE' 'atomic_fetch_orMODE', 'atomic_fetch_andMODE' 'atomic_fetch_xorMODE', 'atomic_fetch_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics, and return the original value. Operand 0 is an output operand which contains the value of the memory location before the operation was performed. Operand 1 is the memory on which the atomic operation is performed. Operand 2 is the second operand to the binary operator. Operand 3 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:%w0 = atomic_fetch_add((<smop> *)%1, %w0)- When compiling the program B GCC decides to emit an `atomic_addDI' insn, documented as:
'
atomic_addMODE', 'atomic_subMODE'
'atomic_orMODE', 'atomic_andMODE'
'atomic_xorMODE', 'atomic_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics. Operand 0 is the memory on which the atomic operation is performed. Operand 1 is the second operand to the binary operator. Operand 2 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:lock *(<smop> *)%w0 += %w1This is done for all targets. In x86-64, for example, case A compiles to:
0000000000000000 <func>: 0: b8 01 00 00 00 mov $0x1,%eax 5: f0 48 0f c1 05 00 00 lock xadd %rax,0x0(%rip) # e <func+0xe> c: 00 00 e: c3 retqAnd case B compiles to:
Quoted 8 lines
0000000000000000 <func>: 0: f0 48 83 05 00 00 00 lock addq $0x1,0x0(%rip) # 9 <func+0x9> 7: 00 01 9: 48 8b 05 00 00 00 00 mov 0x0(%rip),%rax # 10 <func+0x10> 10: c3 retqWhy wouldn't the compiler be allowed to optimize from
atomic_fetch_addtoatomic_addin this case?Ok I see. The generic compiler optimization is ok. It is the backend that is buggy because it emits BPF instruction sequences with different memory ordering semantics for
atomic_OPandatomic_fetch_OP.The only difference between fetching and non-fetching builtins is that in one case the original value is returned, in the other the new value. Other than that they should be equivalent.
For ARM64, GCC generates for case A:
0000000000000000 <func>: 0: 90000001 adrp x1, 0 <func> 4: d2800020 mov x0, #0x1 // #1 8: 91000021 add x1, x1, #0x0 c: f8e00020 ldaddal x0, x0, [x1] 10: d65f03c0 retAnd this for case B:
0000000000000000 <func>: 0: 90000000 adrp x0, 0 <func> 4: d2800022 mov x2, #0x1 // #1 8: 91000001 add x1, x0, #0x0 c: f8e20021 ldaddal x2, x1, [x1] 10: f9400000 ldr x0, [x0] 14: d65f03c0 reti.e. GCC emits LDADDAL for both
atomic_addandatomic_fetch_addinternal insns. Like in x86-64, both sequences have same memory ordering semantics.
Quoted 6 lines
Allright we are changing GCC to always emit fetch versions of sequences for all the supported atomic operations: add, and, or, xor. After the change the `lock' versions of the instructions will not be generated by the compiler at all out of inline asm.
Will send a headsup when done.
Thanks! https://github.com/llvm/llvm-project/pull/101428
is the change in llvm side.
[...]
Quoted 289 lines · Jose E. Marchesi
Quoted 287 lines
Quoted 243 lines
Quoted 152 lines · Alexei Starovoitov
Quoted 125 lines · Peilin Ye
Quoted 26 lines
Hi list!
As we are looking at running sched_ext-style BPF scheduling on architectures with a more relaxed memory model (i.e. ARM), we would like to:
1. have fine-grained control over memory ordering in BPF (instead of defaulting to a full barrier), for performance reasons
2. pay closer attention to if memory barriers are being used correctly in BPFTo that end, our main goal here is to support more types of memory barriers in BPF. While Paul E. McKenney et al. are working on the formalized BPF memory model [1], Paul agreed that it makes sense to support some basic types first. Additionally, we noticed an issue with the __sync_*fetch*() compiler built-ins related to memory ordering, which will be described in details below.
I. We need more types of BPF memory barriers
--------------------------------------------Currently, when it comes to BPF memory barriers, our choices are effectively limited to:
* compiler barrier: 'asm volatile ("" ::: "memory");'
* full memory barriers implied by compiler built-ins like__sync_val_compare_and_swap()Quoted 5 lines
We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.Quoted 71 lines
II. Implicit barriers can get confusing
---------------------------------------We noticed that, as a bit of a surprise, the __sync_*fetch*() built-ins do not always imply a full barrier for BPF on ARM. For example, when using LLVM, the frequently-used
__sync_fetch_and_add()can either imply "relaxed" (no barrier), or "acquire and release" (full barrier) semantics, depending on if its return value is used:Case (a): return value is used
SEC("...") int64_t foo; int64_t func(...) { return __sync_fetch_and_add(&foo, 1); }For case (a), Clang gave us:
3: db 01 00 00 01 00 00 00 r0 = atomic_fetch_add((u64 *)(r1 + 0x0), r0) opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000001): BPF_ADD | BPF_FETCHCase (b): return value is ignored
SEC("...") int64_t foo; int64_t func(...) { __sync_fetch_and_add(&foo, 1); return foo; }For case (b), Clang gave us:
3: db 12 00 00 00 00 00 00 lock *(u64 *)(r2 + 0x0) += r1 opcode (0xdb): BPF_STX | BPF_ATOMIC | BPF_DW imm (0x00000000): BPF_ADDLLVM decided to drop
BPF_FETCH, since the return value of__sync_fetch_and_add()is being ignored [2]. Now, if we take a look atemit_lse_atomic()in the BPF JIT compiler code for ARM64 (suppose that LSE atomic instructions are being used):case BPF_ADD: emit(A64_STADD(isdw, reg, src), ctx); break; <...> case BPF_ADD | BPF_FETCH: emit(A64_LDADDAL(isdw, src, reg, src), ctx); break;STADD is an alias for LDADD. According to [3]:
* LDADDAL for case (a) has "acquire" plus "release" semantics
* LDADD for case (b) "has neither acquire nor release semantics"This is pretty non-intuitive; a compiler built-in should not have inconsistent implications on memory ordering, and it is better not to require all BPF programmers to memorize this.
GCC seems a bit ambiguous [4] on whether __sync_*fetch*() built-ins should always imply a full barrier. GCC considers these __sync_*() built-ins as "legacy", and introduced a new set of __atomic_*() built-ins ("Memory Model Aware Atomic Operations") [5] to replace them. These __atomic_*() built-ins are designed to be a lot more explicit on memory ordering, for example:
Quoted 7 lines
type
__atomic_fetch_add(type *ptr, type val, int memorder)This requires the programmer to specify a memory order type (relaxed, acquire, release...) via the "memorder" parameter. Currently in LLVM, for BPF, those __atomic_*fetch*() built-ins seem to be aliases to their __sync_*fetch*() counterparts (the "memorder" parameter seems effectively ignored), and are not fully supported.
Quoted 5 lines
This sounds like a compiler bug.
Yonghong, Jose,
do you know what compilers do for other backends?
Is it allowed to convertsycn_fetch_addintosync_addwhen fetch part is unused?This behavior is introduced by the following llvm commit:
https://github.com/llvm/llvm-project/commit/286daafd65129228e08a1d07aa4ca74488615744Specifically the following commit message:
=======
Similar to xadd, atomic xadd, xor and xxor (atomic_<op>) instructions are added for atomic operations which do not have return values. LLVM will check the return value for __sync_fetch_and_{add,and,or,xor}.
If the return value is used, instructions atomic_fetch_<op> will be used. Otherwise, atomic_<op> instructions will be used.
======Basically, if no return value,
__sync_fetch_and_add()will use
xadd insn. The decision is made at that time to maintain backward compatibility. For one example, in bcc
https://github.com/iovisor/bcc/blob/master/src/cc/export/helpers.h#L1444
we have
#definelock_xadd(ptr, val)((void)__sync_fetch_and_add(ptr, val))Quoted 3 lines
Should we use atomic_fetch_*() always regardless of whether the return val is used or not? Probably, it should still work. Not sure what gcc does for this case.
GCC behaves similarly.
For program A:
long foo; long func () { return __sync_fetch_and_add(&foo, 1); }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>: 0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 01 00 00 00 r1=atomic_fetch_add((u64*)(r0+0),r1) 20: bf 10 00 00 00 00 00 00 r0=r1 28: 95 00 00 00 00 00 00 00 exitAnd for program B:
long foo; long func () { __sync_fetch_and_add(&foo, 1); return foo; }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>: 0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 00 00 00 00 lock *(u64*)(r0+0)+=r1 20: 79 00 00 00 00 00 00 00 r0=*(u64*)(r0+0) 28: 95 00 00 00 00 00 00 00 exitInternally:
- When compiling the program A GCC decides to emit an `atomic_fetch_addDI' insn, documented as:
'
atomic_fetch_addMODE', 'atomic_fetch_subMODE' 'atomic_fetch_orMODE', 'atomic_fetch_andMODE' 'atomic_fetch_xorMODE', 'atomic_fetch_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics, and return the original value. Operand 0 is an output operand which contains the value of the memory location before the operation was performed. Operand 1 is the memory on which the atomic operation is performed. Operand 2 is the second operand to the binary operator. Operand 3 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:%w0 = atomic_fetch_add((<smop> *)%1, %w0)- When compiling the program B GCC decides to emit an `atomic_addDI' insn, documented as:
'
atomic_addMODE', 'atomic_subMODE'
'atomic_orMODE', 'atomic_andMODE'
'atomic_xorMODE', 'atomic_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics. Operand 0 is the memory on which the atomic operation is performed. Operand 1 is the second operand to the binary operator. Operand 2 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:lock *(<smop> *)%w0 += %w1This is done for all targets. In x86-64, for example, case A compiles to:
0000000000000000 <func>: 0: b8 01 00 00 00 mov $0x1,%eax 5: f0 48 0f c1 05 00 00 lock xadd %rax,0x0(%rip) # e <func+0xe> c: 00 00 e: c3 retqAnd case B compiles to:
Quoted 8 lines
0000000000000000 <func>: 0: f0 48 83 05 00 00 00 lock addq $0x1,0x0(%rip) # 9 <func+0x9> 7: 00 01 9: 48 8b 05 00 00 00 00 mov 0x0(%rip),%rax # 10 <func+0x10> 10: c3 retqWhy wouldn't the compiler be allowed to optimize from
atomic_fetch_addtoatomic_addin this case?Ok I see. The generic compiler optimization is ok. It is the backend that is buggy because it emits BPF instruction sequences with different memory ordering semantics for
atomic_OPandatomic_fetch_OP.The only difference between fetching and non-fetching builtins is that in one case the original value is returned, in the other the new value. Other than that they should be equivalent.
For ARM64, GCC generates for case A:
0000000000000000 <func>: 0: 90000001 adrp x1, 0 <func> 4: d2800020 mov x0, #0x1 // #1 8: 91000021 add x1, x1, #0x0 c: f8e00020 ldaddal x0, x0, [x1] 10: d65f03c0 retAnd this for case B:
0000000000000000 <func>: 0: 90000000 adrp x0, 0 <func> 4: d2800022 mov x2, #0x1 // #1 8: 91000001 add x1, x0, #0x0 c: f8e20021 ldaddal x2, x1, [x1] 10: f9400000 ldr x0, [x0] 14: d65f03c0 reti.e. GCC emits LDADDAL for both
atomic_addandatomic_fetch_addinternal insns. Like in x86-64, both sequences have same memory ordering semantics.Allright we are changing GCC to always emit fetch versions of sequences for all the supported atomic operations: add, and, or, xor. After the change the `lock' versions of the instructions will not be generated by the compiler at all out of inline asm.
Quoted 4 lines
Quoted 1 line
Will send a headsup when done.
Thanks! https://github.com/llvm/llvm-project/pull/101428
is the change in llvm side.
This is now fixed in GCC upstream as well.
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/659454.html
Salud!
Hi Jose,
Quoted 130 lines · Jose E. Marchesi
Quoted 90 lines
GCC behaves similarly.
For program A:
long foo;
long func () { return __sync_fetch_and_add(&foo, 1); }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>:
0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 01 00 00 00 r1=atomic_fetch_add((u64*)(r0+0),r1) 20: bf 10 00 00 00 00 00 00 r0=r1 28: 95 00 00 00 00 00 00 00 exitAnd for program B:
long foo;
long func () { __sync_fetch_and_add(&foo, 1); return foo; }bpf-unknown-none-gcc -O2 compiles to:
0000000000000000 <func>:
0: 18 00 00 00 00 00 00 00 r0=0 ll 8: 00 00 00 00 00 00 00 00 10: b7 01 00 00 01 00 00 00 r1=1 18: db 10 00 00 00 00 00 00 lock *(u64*)(r0+0)+=r1 20: 79 00 00 00 00 00 00 00 r0=*(u64*)(r0+0) 28: 95 00 00 00 00 00 00 00 exitInternally:
- When compiling the program A GCC decides to emit an `atomic_fetch_addDI' insn, documented as:
'
atomic_fetch_addMODE', 'atomic_fetch_subMODE'
'atomic_fetch_orMODE', 'atomic_fetch_andMODE'
'atomic_fetch_xorMODE', 'atomic_fetch_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics, and return the original value. Operand 0 is an output operand which contains the value of the memory location before the operation was performed. Operand 1 is the memory on which the atomic operation is performed. Operand 2 is the second operand to the binary operator. Operand 3 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:%w0 = atomic_fetch_add((<smop> *)%1, %w0)- When compiling the program B GCC decides to emit an `atomic_addDI' insn, documented as:
'
atomic_addMODE', 'atomic_subMODE'
'atomic_orMODE', 'atomic_andMODE'
'atomic_xorMODE', 'atomic_nandMODE'These patterns emit code for an atomic operation on memory with memory model semantics. Operand 0 is the memory on which the atomic operation is performed. Operand 1 is the second operand to the binary operator. Operand 2 is the memory model to be used by the operation.
The BPF backend defines
atomic_fetch_addfor DI modes (long) to expand to this BPF instruction:lock *(<smop> *)%w0 += %w1This is done for all targets. In x86-64, for example, case A compiles to:
0000000000000000 <func>:
0: b8 01 00 00 00 mov $0x1,%eax 5: f0 48 0f c1 05 00 00 lock xadd %rax,0x0(%rip) # e <func+0xe> c: 00 00 e: c3 retqAnd case B compiles to:
Quoted 8 lines
0000000000000000 <func>:
0: f0 48 83 05 00 00 00 lock addq $0x1,0x0(%rip) # 9 <func+0x9> 7: 00 01 9: 48 8b 05 00 00 00 00 mov 0x0(%rip),%rax # 10 <func+0x10> 10: c3 retqWhy wouldn't the compiler be allowed to optimize from
atomic_fetch_addtoatomic_addin this case?Ok I see. The generic compiler optimization is ok. It is the backend that is buggy because it emits BPF instruction sequences with different memory ordering semantics for
atomic_OPandatomic_fetch_OP.The only difference between fetching and non-fetching builtins is that in one case the original value is returned, in the other the new value. Other than that they should be equivalent.
For ARM64, GCC generates for case A:
0000000000000000 <func>:
0: 90000001 adrp x1, 0 <func> 4: d2800020 mov x0, #0x1 // #1 8: 91000021 add x1, x1, #0x0 c: f8e00020 ldaddal x0, x0, [x1] 10: d65f03c0 retAnd this for case B:
0000000000000000 <func>:
0: 90000000 adrp x0, 0 <func> 4: d2800022 mov x2, #0x1 // #1 8: 91000001 add x1, x0, #0x0 c: f8e20021 ldaddal x2, x1, [x1] 10: f9400000 ldr x0, [x0] 14: d65f03c0 reti.e. GCC emits LDADDAL for both
atomic_addandatomic_fetch_addinternal insns. Like in x86-64, both sequences have same memory ordering semantics.
Quoted 6 lines
Allright we are changing GCC to always emit fetch versions of sequences for all the supported atomic operations: add, and, or, xor. After the change the `lock' versions of the instructions will not be generated by the compiler at all out of inline asm.
Will send a headsup when done.
Thanks for taking care of this!
Peilin Ye
Hi Alexei,
Thanks for all the suggestions! Some questions:
Quoted 9 lines · Alexei Starovoitov
Quoted 5 lines · Peilin Ye
We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.
I see, so something like:
asm volatile("%0 = load_acquire((u64 *)(%1 + 0x0))" :
"=r"(ret) : "r"(&foo) : "memory");and e.g. this in disassembly:
r0 = load_acquire((u64 *)(r1 + 0x0))
I agree that we'd better not put the entire e.g.
"r0 = __atomic_load_n((u64 *)(r1 + 0x0), __ATOMIC_ACQUIRE)" into disassembly.
Quoted 12 lines
Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.
[...]
Quoted 7 lines
Quoted 4 lines
Roughly, the scope of this work includes:
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?
Do we allow aliases? E.g., can we have "MEMACQ" for LDX and "MEMREL" for STX, but let them share the same numeric value?
Speaking of numeric value, out of curiosity:
IMM 0
ABS 1
IND 2
MEM 3
MEMSX 4ATOMIC 6
Was there a reason that we skipped 5? Is 5 reserved?
Quoted 10 lines
For barriers we would need a new insn. Not sure which class would fit the best. Maybe
BPF_LD?Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel codesmp_mb(),smp_rmb(), etc.
Quoted 2 lines
I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Ah, I see; we recently supported [1] nocsr BPF helper functions. The cover letter says:
"""
This patch-set seeks to allow using no_caller_saved_registers gcc/clang attribute with some BPF helper functions (and kfuncs in the future).
"""
It seems that nocsr BPF kfuncs are not supported yet. Do we have a schedule for it?
Thanks,
Peilin Ye
[1] [PATCH bpf-next v4 00/10] no_caller_saved_registers attribute for helper calls
https://lore.kernel.org/bpf/[email protected]/
Quoted 16 lines · Peilin Ye
Hi Alexei,
Thanks for all the suggestions! Some questions:
Quoted 9 lines · Alexei Starovoitov
Quoted 5 lines · Peilin Ye
We need more. During offline discussion with Paul, we agreed we can start from:
* load-acquire: __atomic_load_n(... memorder=__ATOMIC_ACQUIRE);
* store-release: __atomic_store_n(... memorder=__ATOMIC_RELEASE);we would need inline asm equivalent too. Similar to kernel
smp_load_acquire()macro.
Quoted 8 lines
I see, so something like:
asm volatile("%0 = load_acquire((u64 *)(%1 + 0x0))" : "=r"(ret) : "r"(&foo) : "memory");and e.g. this in disassembly:
r0 = load_acquire((u64 *)(r1 + 0x0))
yes.
Quoted 3 lines
I agree that we'd better not put the entire e.g.
"r0 =__atomic_load_n((u64 *)(r1 + 0x0), __ATOMIC_ACQUIRE)" into disassembly.
exactly. That's too verbose.
Quoted 17 lines
Quoted 12 lines
Quoted 5 lines
Theoretically, the BPF JIT compiler could also reorder instructions just like Clang or GCC, though it might not currently do so. If we ever developed a more optimizing BPF JIT compiler, it would also be nice to have an optimization barrier for it. However, Alexei Starovoitov has expressed that defining a BPF instruction with 'asm volatile ("" ::: "memory");' semantics might be tricky.
It can be a standalone insn that is a compiler barrier only but that feels like a waste of an instruction. So depending how we end up encoding various real barriers
there may be a bit to spend in such a barrier insn that is only a compiler barrier.
In this case optimizing JIT barrier.[...]
Quoted 2 lines
Quoted 2 lines
Roughly, the scope of this work includes:
Quoted 8 lines
Quoted 5 lines
Quoted 2 lines
* decide how to extend the BPF ISA (add new instructions and/or extend current ones)
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?Do we allow aliases? E.g., can we have "MEMACQ" for LDX and "MEMREL" for STX, but let them share the same numeric value?
yes. See
#define BPF_ATOMIC 0xc0 /* atomic memory ops - op type in immediate */ #define BPF_XADD 0xc0 /* exclusive add - legacy name */
but it has to be backward compatible.
Quoted 10 lines
Speaking of numeric value, out of curiosity:
IMM 0 ABS 1 IND 2 MEM 3 MEMSX 4ATOMIC 6
Was there a reason that we skipped 5? Is 5 reserved?
See
/* unused opcode to mark special load instruction. Same as BPF_ABS */
#define BPF_PROBE_MEM 0x20
/* unused opcode to mark special ldsx instruction. Same as BPF_IND */ #define BPF_PROBE_MEMSX 0x40
/* unused opcode to mark special load instruction. Same as BPF_MSH */ #define BPF_PROBE_MEM32 0xa0
it's used by the verifier when it remaps opcode to tell JIT. It can be used, but then the internal opcode needs to change too.
Quoted 16 lines
Quoted 10 lines
For barriers we would need a new insn. Not sure which class would fit the best. Maybe
BPF_LD?Another alternative for barriers is to use nocsr kfuncs. Then we have the freedom to make mistakes and fix them later. One kfunc per barrier would do.
JITs would inline them into appropriate insns.
In bpf progs they will be used just like in the kernel codesmp_mb(),smp_rmb(), etc.Quoted 2 lines
I don't think compilers have to emit barriers from C code, so my preference is kfuncs atm.
Ah, I see; we recently supported [1] nocsr BPF helper functions. The cover letter says:
Quoted 8 lines
"""
This patch-set seeks to allow usingno_caller_saved_registersgcc/clang attribute with some BPF helper functions (and kfuncs in the future).
"""It seems that nocsr BPF kfuncs are not supported yet. Do we have a schedule for it?
Support for nocsr for kfuncs is being added.
Assume it's already available :)
It's not a blocker to add barrier kfuncs.
Quoted 35 lines · Alexei Starovoitov
Quoted 5 lines
Quoted 2 lines
ldx/stx insns support MEM and MEMSX modifiers. Adding
MEM_ACQ_RELfeels like a natural fit. Better name?Do we allow aliases? E.g., can we have "MEMACQ" for LDX and "MEMREL" for STX, but let them share the same numeric value?
yes. See
#define BPF_ATOMIC 0xc0 /* atomic memory ops - op type in immediate */ #define BPF_XADD 0xc0 /* exclusive add - legacy name */but it has to be backward compatible.
Quoted 10 lines
Speaking of numeric value, out of curiosity:
IMM 0 ABS 1 IND 2 MEM 3 MEMSX 4ATOMIC 6
Was there a reason that we skipped 5? Is 5 reserved?
See
/* unused opcode to mark special load instruction. Same asBPF_ABS*/
#defineBPF_PROBE_MEM0x20/* unused opcode to mark special ldsx instruction. Same as
BPF_IND*/ #defineBPF_PROBE_MEMSX0x40/* unused opcode to mark special load instruction. Same as
BPF_MSH*/ #defineBPF_PROBE_MEM320xa0it's used by the verifier when it remaps opcode to tell JIT. It can be used, but then the internal opcode needs to change too.
[...]
Quoted 6 lines
Quoted 2 lines
It seems that nocsr BPF kfuncs are not supported yet. Do we have a schedule for it?
Support for nocsr for kfuncs is being added.
Assume it's already available :)
It's not a blocker to add barrier kfuncs.
Got it! I'll start cooking (kernel and LLVM) patches for "MEMACQ" and "MEMREL" (using 0x7) first.
Thanks,
Peilin Ye
Hi all,
LLVM changes for load-acquire and store-release are available for review
at: https://github.com/llvm/llvm-project/pull/108636
I've tentatively put them under -mcpu=v5. Please take a look when you have a chance. Thanks!
I just noticed, however:
Quoted 16 lines · Alexei Starovoitov
Quoted 15 lines
Quoted 10 lines
Speaking of numeric value, out of curiosity:
IMM 0 ABS 1 IND 2 MEM 3 MEMSX 4ATOMIC 6
Was there a reason that we skipped 5? Is 5 reserved?
See
/* unused opcode to mark special load instruction. Same asBPF_ABS*/
#defineBPF_PROBE_MEM0x20
Quoted 8 lines
Quoted 8 lines
/* unused opcode to mark special ldsx instruction. Same as
BPF_IND*/ #defineBPF_PROBE_MEMSX0x40/* unused opcode to mark special load instruction. Same as
BPF_MSH*/ #defineBPF_PROBE_MEM320xa0it's used by the verifier when it remaps opcode to tell JIT. It can be used, but then the internal opcode needs to change too.
There's also:
/* unused opcode to mark special atomic instruction */ #define BPF_PROBE_ATOMIC 0xe0
0xe0 is (7 << 5), so it seems like we've already run out of bits for mode modifiers? Can we delete these internal opcodes and re-implement them in other ways, to make room for MEMACQ (MEMREL)?
Thanks,
Peilin Ye
Quoted 27 lines · Peilin Ye
I just noticed, however:
Quoted 16 lines · Alexei Starovoitov
Quoted 15 lines
Quoted 10 lines
Speaking of numeric value, out of curiosity:
IMM 0 ABS 1 IND 2 MEM 3 MEMSX 4ATOMIC 6
Was there a reason that we skipped 5? Is 5 reserved?
See
/* unused opcode to mark special load instruction. Same asBPF_ABS*/
#defineBPF_PROBE_MEM0x20Quoted 8 lines
Quoted 8 lines
/* unused opcode to mark special ldsx instruction. Same as
BPF_IND*/ #defineBPF_PROBE_MEMSX0x40/* unused opcode to mark special load instruction. Same as
BPF_MSH*/ #defineBPF_PROBE_MEM320xa0it's used by the verifier when it remaps opcode to tell JIT. It can be used, but then the internal opcode needs to change too.
Quoted 8 lines
There's also:
/* unused opcode to mark special atomic instruction */ #define
BPF_PROBE_ATOMIC0xe0
0xe0is (7 << 5), so it seems like we've already run out of bits for mode modifiers? Can we delete these internal opcodes and re-implement them in other ways, to make room for MEMACQ (MEMREL)?
Can we instead use the unused higher bits of bpf_insn::imm as a scratch
pad? ^^^
For example, instead of letting the verifier change bpf_insn::code from BPF_ATOMIC to BPF_PROBE_ATOMIC, keep it as-is, but set the highest bit in bpf_insn::imm to let the JIT know it is a "_PROBE_" instruction, and there's a corresponding entry in the exception table.
Thanks,
Peilin Ye