Support riscv jit to provide
Subsystems (7)
Pu Lehui
maintainerv2 cover Support riscv jit to provide
patch 1 fix an issue that could not print bpf line info due to data inconsistency in 32-bit environment.
patch 2 add support for riscv jit to provide bpf_line_info. Both RV32 and RV64 tests have been passed as like follow:
./test_progs -a btf
#19 btf:OK
Summary: 1/215 PASSED, 0 SKIPPED, 0 FAILED
v2:
- Remove some trivial code
v1: https://lore.kernel.org/bpf/[email protected]
Pu Lehui (2):
bpf: Unify data extension operation of jited_ksyms and jited_linfo riscv, bpf: Support riscv jit to provide bpf_line_info
arch/riscv/net/bpf_jit.h | 1 + arch/riscv/net/bpf_jit_core.c | 7 ++++++- kernel/bpf/syscall.c | 5 ++++- 3 files changed, 11 insertions(+), 2 deletions(-)
Signature · 1 line
2.25.1
Pu Lehui
maintainerPATCH v2 1/2 bpf: Unify data extension operation of jited_ksyms and jited_linfo
We found that 32-bit environment can not print bpf line info due to data inconsistency between jited_ksyms[0] and jited_linfo[0].
For example:
jited_kyms[0] = 0xb800067c, jited_linfo[0] = 0xffffffffb800067c
We know that both of them store bpf func address, but due to the different data extension operations when extended to u64, they may not be the same. We need to unify the data extension operations of them.
--- kernel/bpf/syscall.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
kernel/bpf/syscall.c +4 -1
@@ -3871,13 +3871,16 @@ static int bpf_prog_get_info_by_fd(struct file *file, info.nr_jited_line_info =0 ;if (info.nr_jited_line_info && ulen) {if (bpf_dump_raw_ok (file->f_cred)) { +unsigned long jited_linfo_addr;__u64 __user *user_linfo;u32 i; user_linfo =u64_to_user_ptr (info.jited_line_info); ulen =min_t (u32, info.nr_jited_line_info, ulen);for (i =0 ; i < ulen; i++) { -if (put_user ((__u64 )(long )prog->aux->jited_linfo[i], + jited_linfo_addr = (unsigned long ) + prog->aux->jited_linfo[i]; +if (put_user ((__u64 ) jited_linfo_addr, &user_linfo[i]))return -EFAULT ; }
Signature · 1 line
2.25.1
John Fastabend
reviewerRE: [PATCH bpf-next v2 1/2] bpf: Unify data extension operation of jited_ksyms and jited_linfo
Quoted 26 lines · Pu Lehui
We found that 32-bit environment can not print bpf line info due to data inconsistency between jited_ksyms[0] and jited_linfo[0].
For example:
jited_kyms[0] =0xb800067c, jited_linfo[0] =0xffffffffb800067cWe know that both of them store bpf func address, but due to the different data extension operations when extended to u64, they may not be the same. We need to unify the data extension operations of them.
Signed-off-by Pu Lehui <[email protected]>--- kernel/bpf/syscall.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)kernel/bpf/syscall.c +1
-0@@ -3871,13 +3871,16 @@ static int bpf_prog_get_info_by_fd(struct file *file, info.nr_jited_line_info =0 ;if (info.nr_jited_line_info && ulen) {if (bpf_dump_raw_ok (file->f_cred)) { +unsigned long jited_linfo_addr;__u64 __user *user_linfo;
kernel/bpf/syscall.c static int bpf_prog_get_info_by_fd(struct file *file, · line 3884
Quoted 10 lines
kernel/bpf/syscall.c +3
-1u32 i; user_linfo =u64_to_user_ptr (info.jited_line_info); ulen =min_t (u32, info.nr_jited_line_info, ulen);for (i =0 ; i < ulen; i++) { -if (put_user ((__u64 )(long )prog->aux->jited_linfo[i], + jited_linfo_addr = (unsigned long ) + prog->aux->jited_linfo[i]; +if (put_user ((__u64 ) jited_linfo_addr, &user_linfo[i]))
the logic is fine but i'm going to nitpick a bit this 4 lines is ugly just make it slightly longer than 80chars or use a shoarter name? For example,
for (i = 0; i < ulen; i++) {
unsigned long l;
l = (unsigned long) prog->aux->jited_linfo[i];
if (put_user((__u64) l, &user_linfo[i]))is much nicer -- no reason to smash single assignment across multiple lines. My $.02.
Thanks,
John
Pu Lehui
maintainerRe: [PATCH bpf-next v2 1/2] bpf: Unify data extension operation of jited_ksyms and jited_linfo
Quoted 42 lines · John Fastabend
Quoted 26 lines · Pu Lehui
We found that 32-bit environment can not print bpf line info due to data inconsistency between jited_ksyms[0] and jited_linfo[0].
For example:
jited_kyms[0] =0xb800067c, jited_linfo[0] =0xffffffffb800067cWe know that both of them store bpf func address, but due to the different data extension operations when extended to u64, they may not be the same. We need to unify the data extension operations of them.
Signed-off-by Pu Lehui <[email protected]>--- kernel/bpf/syscall.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)kernel/bpf/syscall.c +1
-0@@ -3871,13 +3871,16 @@ static int bpf_prog_get_info_by_fd(struct file *file, info.nr_jited_line_info =0 ;if (info.nr_jited_line_info && ulen) {if (bpf_dump_raw_ok (file->f_cred)) { +unsigned long jited_linfo_addr;__u64 __user *user_linfo;kernel/bpf/syscall.c static int bpf_prog_get_info_by_fd(struct file *file, · line 3884
Quoted 10 lines
kernel/bpf/syscall.c +3
-1u32 i; user_linfo =u64_to_user_ptr (info.jited_line_info); ulen =min_t (u32, info.nr_jited_line_info, ulen);for (i =0 ; i < ulen; i++) { -if (put_user ((__u64 )(long )prog->aux->jited_linfo[i], + jited_linfo_addr = (unsigned long ) + prog->aux->jited_linfo[i]; +if (put_user ((__u64 ) jited_linfo_addr, &user_linfo[i]))the logic is fine but i'm going to nitpick a bit this 4 lines is ugly just make it slightly longer than 80chars or use a shoarter name? For example,
Quoted 9 lines
for (i = 0; i < ulen; i++) { unsigned long l; l = (unsigned long) prog->aux->jited_linfo[i]; if (put_user((__u64) l, &user_linfo[i]))is much nicer -- no reason to smash single assignment across multiple lines. My $.02.
Okay, It sounds good. I will make change in next version. Thanks.
Quoted 4 lines
Thanks,
John
.
Pu Lehui
maintainerPATCH v2 2/2 riscv, bpf: Support riscv jit to provide bpf_line_info
In reply to Pu LehuiAdd support for riscv jit to provide bpf_line_info. We need to consider the prologue offset in ctx->offset, but unlike x86 and arm64, ctx->offset of riscv does not provide an extra slot for the prologue, so here we just calculate the len of prologue and add it to ctx->offset at the end. Both RV64 and RV32 have been tested.
--- arch/riscv/net/bpf_jit.h | 1 + arch/riscv/net/bpf_jit_core.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletion(-)
arch/riscv/net/bpf_jit.h +1 -0
@@ -69,6 +69,7 @@ struct rv_jit_context {struct bpf_prog *prog;u16 *insns;/* RV insns */ int ninsns; +int prologue_offset;int epilogue_offset;int *offset;/* BPF to RV */ int nexentries;
arch/riscv/net/bpf_jit_core.c +6 -1
@@ -44,7 +44,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)unsigned int prog_size =0 , extable_size =0 ;bool tmp_blinded = false, extra_pass = false;struct bpf_prog *tmp, *orig_prog = prog; -int pass =0 , prev_ninsns =0 , i; +int pass =0 , prev_ninsns =0 , prologue_len, i;struct rv_jit_data *jit_data;struct rv_jit_context *ctx;
@@ -95,6 +95,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog) prog = orig_prog; goto out_offset; } + ctx->prologue_offset = ctx->ninsns;bpf_jit_build_prologue (ctx); ctx->epilogue_offset = ctx->ninsns;bpf_jit_build_epilogue (ctx);
@@ -161,6 +162,10 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)if (!prog->is_func || extra_pass) {bpf_jit_binary_lock_ro (jit_data->header); + prologue_len = ctx->epilogue_offset - ctx->prologue_offset; +for (i =0 ; i < prog->len; i++) + ctx->offset[i] =ninsns_rvoff (prologue_len + ctx->offset[i]); +bpf_prog_fill_jited_linfo (prog, ctx->offset); out_offset:kfree (ctx->offset);kfree (jit_data);
Signature · 1 line
2.25.1
John Fastabend
reviewerRE: [PATCH bpf-next v2 2/2] riscv, bpf: Support riscv jit to provide bpf_line_info
Quoted 9 lines · Pu Lehui
Add support for riscv jit to provide
bpf_line_info. We need to consider the prologue offset inctx->offset, but unlike x86 and arm64,ctx->offsetof riscv does not provide an extra slot for the prologue, so here we just calculate the len of prologue and add it toctx->offsetat the end. Both RV64 and RV32 have been tested.Signed-off-by Pu Lehui <[email protected]>---
Looks reasonable to me, but would be good if someone with RISC knowledge takes a lookt hough.
_______________________________________________
linux-riscv mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-riscv
Luke Nelson
maintainerRe: [PATCH bpf-next v2 2/2] riscv, bpf: Support riscv jit to provide bpf_line_info
In reply to Pu LehuiThanks for the patch! I have a couple of notes written down below.
arch/riscv/net/bpf_jit_core.c struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog) · line 167
Quoted 5 lines
arch/riscv/net/bpf_jit_core.c +4
-0+ ctx->prologue_offset = ctx->ninsns; + prologue_len = ctx->epilogue_offset - ctx->prologue_offset; +for (i =0 ; i < prog->len; i++) + ctx->offset[i] =ninsns_rvoff (prologue_len + ctx->offset[i]);
The logic looks correct to me; my only nit is that the name prologue_offset might be a bit confusing. The prologue is always at the beginning of the final JITed program, it just happens to be that the prologue is emitted "out of order" on the initial/internal passes that compute offsets.
What prologue_offset really measures in your code is the length of the body of the JITed program. What do you think about renaming prologue_offset to something like body_len? Then the line to compute prologue_len becomes:
prologue_len = ctx->epilogue_offset - ctx->body_len;
This version makes more sense to me why it's correct. Curious what you think.
Quoted 1 line
+
bpf_prog_fill_jited_linfo(prog, ctx->offset);
Here's a quote from the comment that documentsbpf_prog_fill_jited_linfo in kernel/bpf/core.c:
/* The jit engine is responsible to provide an array
* for insn_off to the jited_off mapping (insn_to_jit_off).
...
* jited_off is the byte off to the last byte of the jited insn.
This comment says that ctx->offset (passed to this function as insn_to_jit_off) should map each instruction to the offset of the last byte of the JITed instructions, but as I understand it your patch sets ctx->offset[i] to be the offset _one past_ the last byte of the JITed instructions (i.e., the first byte of the next instruction). I'm not sure if this is a bug in your code, in this comment, or in my understanding :)
As a concrete example, suppose the BPF instruction at index 0 compiles to 2 (non-compressed) RISC-V instructions, or 8 bytes. Then ctx->offset[0] will be 2 after the initial JIT passes, and your code would update ctx->offset[0] to be 4*prologue_len + 8. This offset corresponds to the first byte of insns[1], not the last byte of insn[0], which would be 4*prologue_len + 7.
My guess would be that the comment is out of date and your code is doing the correct thing, since it seems in line with what other JITs are doing. If that's the case, maybe we can consider updating that comment at some point. I'm curious if the tests you ran would break if you changed your code to match what the comment says (i.e., subtracting 1 byte from each element in ctx->offset before passing to bpf_prog_fill_jited_linfo).
Quoted 3 lines
./test_progs -a btf
#19 btf:OK
Summary: 1/215 PASSED, 0 SKIPPED, 0 FAILED
Last, did you have a chance to run any of the other tests with your change (e.g., test_verifier, test_bpf.ko, other tests in test_progs)? I don't expect this change to break any tests, but may as well run them if it's easy enough just to be sure.
Thanks!
- Luke
Pu Lehui
maintainerRe: [PATCH bpf-next v2 2/2] riscv, bpf: Support riscv jit to provide bpf_line_info
Hi Luke,
Quoted 14 lines · Luke Nelson
Thanks for the patch! I have a couple of notes written down below.
arch/riscv/net/bpf_jit_core.c struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog) · line 167
Quoted 5 lines
arch/riscv/net/bpf_jit_core.c +4
-0+ ctx->prologue_offset = ctx->ninsns; + prologue_len = ctx->epilogue_offset - ctx->prologue_offset; +for (i =0 ; i < prog->len; i++) + ctx->offset[i] =ninsns_rvoff (prologue_len + ctx->offset[i]);The logic looks correct to me; my only nit is that the name
prologue_offsetmight be a bit confusing. The prologue is always at the beginning of the final JITed program, it just happens to be that the prologue is emitted "out of order" on the initial/internal passes that compute offsets.
Quoted 9 lines
What
prologue_offsetreally measures in your code is the length of the body of the JITed program. What do you think about renamingprologue_offsetto something likebody_len? Then the line to computeprologue_lenbecomes:prologue_len = ctx->epilogue_offset - ctx->body_len;This version makes more sense to me why it's correct. Curious what you think.
Sorry for getting back to you so late. Thanks so much for your review. It seems that ctx->body_len makes more sence, I will rename it.
Quoted 26 lines
Quoted 2 lines
+
bpf_prog_fill_jited_linfo(prog, ctx->offset);Here's a quote from the comment that documents
bpf_prog_fill_jited_linfoinkernel/bpf/core.c:/* The jit engine is responsible to provide an array
* forinsn_offto thejited_offmapping (insn_to_jit_off).
...
*jited_offis the byte off to the last byte of the jited insn.This comment says that
ctx->offset(passed to this function asinsn_to_jit_off) should map each instruction to the offset of the last byte of the JITed instructions, but as I understand it your patch sets ctx->offset[i] to be the offset_onepast_the last byte of the JITed instructions (i.e., the first byte of the next instruction). I'm not sure if this is a bug in your code, in this comment, or in my understanding :)As a concrete example, suppose the BPF instruction at index 0 compiles to 2 (non-compressed) RISC-V instructions, or 8 bytes. Then ctx->offset[0] will be 2 after the initial JIT passes, and your code would update ctx->offset[0] to be 4*prologue_len + 8. This offset corresponds to the first byte of insns[1], not the last byte of insn[0], which would be 4*prologue_len + 7.
Quoted 8 lines
My guess would be that the comment is out of date and your code is doing the correct thing, since it seems in line with what other JITs are doing. If that's the case, maybe we can consider updating that comment at some point. I'm curious if the tests you ran would break if you changed your code to match what the comment says (i.e., subtracting 1 byte from each element in
ctx->offsetbefore passing tobpf_prog_fill_jited_linfo).
IIUC,ctx->offset(passed to bpf_prog_fill_jited_linfo as insn_to_jit_off)
should be the first byte of the next instruction, or the byte off to the end of the current instruction.
Here's the code as belowbpf_prog_fill_jited_linfo in kernel/bpf/core.c:
jited_linfo[i] = prog->bpf_func + insn_to_jit_off[linfo[i].insn_off - insn_start - 1];
we can see here that "linfo[i].insn_off - insn_start - 1" refers to the previous instruction, and the corresponding insn_to_jit_off refers to the first byte of the current instruction.
It seems the following quote might make more sense bpf_prog_fill_jited_linfo in kernel/bpf/core.c:
* jited_off is the byte off to the "end" of the jited insn.
Quoted 10 lines
Quoted 4 lines
./test_progs -a btf
#19 btf:OK
Summary: 1/215 PASSED, 0 SKIPPED, 0 FAILEDLast, did you have a chance to run any of the other tests with your change (e.g.,
test_verifier, test_bpf.ko, other tests intest_progs)? I don't expect this change to break any tests, but may as well run them if it's easy enough just to be sure.
Yeah, "test_verifier", "test_bpf.ko" and "test_progs -a btf" all test pass, as well as "test_progs" with no new failure ceses. I will attach the test result in v3.
Thanks,
Lehui
Quoted 5 lines
Thanks!
- Luke
.