mbox series

[V11,0/5] riscv: Optimize function trace

Message ID 20230627111612.761164-1-suagrfillet@gmail.com (mailing list archive)
Headers show
Series riscv: Optimize function trace | expand

Message

Song Shuai June 27, 2023, 11:16 a.m. UTC
Changes in V11:

- append a patch that makes the DIRECT_CALL samples support RV32I in
  this series fixing the rv32 build failure reported by Palmer

- validated with ftrace boottime selftest and manual sample modules test
  in qemu-system for RV32I and RV64I

This series optimizes function trace. The first 3 independent 
patches has been picked in the V7 version of this series, the
subsequent version continues the following 4 patches:

select FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY [1] (patch 1)
==========================================================

In RISC-V, -fpatchable-function-entry option is used to support
dynamic ftrace in this commit afc76b8b8011 ("riscv: Using
PATCHABLE_FUNCTION_ENTRY instead of MCOUNT"). So recordmcount
don't have to be called to create the __mcount_loc section before
the vmlinux linking.

Here selects FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY to tell
Makefile not to run recordmcount.

Make function graph use ftrace directly [2] (patch 2)
======================================================== 

In RISC-V architecture, when we enable the ftrace_graph tracer on some
functions, the function tracings on other functions will suffer extra
graph tracing work. In essence, graph_ops isn't limited by its func_hash
due to the global ftrace_graph_[regs]_call label. That should be
corrected.

What inspires me is the commit 0c0593b45c9b ("x86/ftrace: Make function
graph use ftrace directly") that uses graph_ops::func function to
install return_hooker and makes the function called against its
func_hash.

Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
==============================================

This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS support for RISC-V.
SAMPLE_FTRACE_DIRECT and SAMPLE_FTRACE_DIRECT_MULTI are also included
here as the samples for testing DIRECT_CALLS related interface.

First, select the DYNAMIC_FTRACE_WITH_DIRECT_CALLS to provide 
register_ftrace_direct[_multi] interfaces allowing user to register 
the customed trampoline (direct_caller) as the mcount for one or 
more target functions. And modify_ftrace_direct[_multi] are also 
provided for modify direct_caller.

At the same time, the samples in ./samples/ftrace/ can be built
as kerenl module for testing these interfaces with SAMPLE_FTRACE_DIRECT
and SAMPLE_FTRACE_DIRECT_MULTI selected.

Second, to make the direct_caller and the other ftrace hooks
(eg. function/fgraph tracer, k[ret]probes) co-exist, a temporary
register
are nominated to store the address of direct_caller in
ftrace_regs_caller.
After the setting of the address direct_caller by direct_ops->func and
the RESTORE_REGS in ftrace_regs_caller, direct_caller will be jumped to
by the `jr` inst.

The series's old changes related these patches
==========================================

Changes in v10:
https://lore.kernel.org/all/20230511093234.3123181-1-suagrfillet@gmail.com/

- add Acked-by from Björn Töpel in patch 2 and patch 4 
- replace `move` with `mv` in patch3 
- prettify patch 2/4 with proper tabs

Changes in v9:
https://lore.kernel.org/linux-riscv/20230510101857.2953955-1-suagrfillet@gmail.com/

1. add Acked-by from Björn Töpel in patch 1

2. rebase patch2/patch3 on Linux v6.4-rc1

  - patch 2: to make the `SAVE_ABI_REGS` configurable, revert the
    modification of mcount-dyn.S from commit (45b32b946a97 "riscv:
entry: Consolidate general regs saving/restoring")

  - patch 3: to pass the trace_selftest, add the implement of
    `ftrace_stub_direct_tramp` from commit (fee86a4ed536 "ftrace:
selftest: remove broken trace_direct_tramp") ; and fixup the context
conflict in Kconfig 

Changes in v8:
https://lore.kernel.org/linux-riscv/20230324033342.3177979-1-suagrfillet@gmail.com/
 - Fix incorrect address values in the 4nd patch 
 - Rebased on v6.3-rc2

Changes in v7:
https://lore.kernel.org/linux-riscv/20230112090603.1295340-1-guoren@kernel.org/
 - Fixup RESTORE_ABI_REGS by remove PT_T0(sp) overwrite.
 - Add FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY [1]
 - Fixup kconfig with HAVE_SAMPLE_FTRACE_DIRECT &
   HAVE_SAMPLE_FTRACE_DIRECT_MULTI

Changes in v6:
https://lore.kernel.org/linux-riscv/20230107133549.4192639-1-guoren@kernel.org/
 - Replace 8 with MCOUNT_INSN_SIZE
 - Replace "REG_L a1, PT_RA(sp)" with "mv a1, ra"
 - Add Evgenii Shatokhin comment

Changes in v5:
https://lore.kernel.org/linux-riscv/20221208091244.203407-1-guoren@kernel.org/
 - Sort Kconfig entries in alphabetical order.

Changes in v4:
https://lore.kernel.org/linux-riscv/20221129033230.255947-1-guoren@kernel.org/
 - Include [3] for maintenance. [Song Shuai]

Changes in V3:
https://lore.kernel.org/linux-riscv/20221123153950.2911981-1-guoren@kernel.org/
 - Include [2] for maintenance. [Song Shuai]

[1]: https://lore.kernel.org/linux-riscv/CAAYs2=j3Eak9vU6xbAw0zPuoh00rh8v5C2U3fePkokZFibWs2g@mail.gmail.com/T/#t
[2]: https://lore.kernel.org/lkml/20221120084230.910152-1-suagrfillet@gmail.com/
[3]: https://lore.kernel.org/linux-riscv/20221123142025.1504030-1-suagrfillet@gmail.com/ 

Song Shuai (5):
  riscv: select FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
  riscv: ftrace: Add ftrace_graph_func
  riscv: ftrace: Add DYNAMIC_FTRACE_WITH_DIRECT_CALLS support
  samples: ftrace: Add riscv support for SAMPLE_FTRACE_DIRECT[_MULTI]
  samples: ftrace: Make the riscv samples support RV32I

 arch/riscv/Kconfig                          |   4 +
 arch/riscv/include/asm/ftrace.h             |  19 +-
 arch/riscv/kernel/ftrace.c                  |  30 ++-
 arch/riscv/kernel/mcount-dyn.S              | 200 ++++++++++++++++----
 samples/ftrace/ftrace-direct-modify.c       |  35 ++++
 samples/ftrace/ftrace-direct-multi-modify.c |  41 ++++
 samples/ftrace/ftrace-direct-multi.c        |  25 +++
 samples/ftrace/ftrace-direct-too.c          |  28 +++
 samples/ftrace/ftrace-direct.c              |  24 +++
 9 files changed, 350 insertions(+), 56 deletions(-)

Comments

Song Shuai July 6, 2023, 9:35 a.m. UTC | #1
Ping...

在 2023/6/27 19:16, Song Shuai 写道:
> Changes in V11:
> 
> - append a patch that makes the DIRECT_CALL samples support RV32I in
>    this series fixing the rv32 build failure reported by Palmer
> 
> - validated with ftrace boottime selftest and manual sample modules test
>    in qemu-system for RV32I and RV64I
> 
> This series optimizes function trace. The first 3 independent
> patches has been picked in the V7 version of this series, the
> subsequent version continues the following 4 patches:
> 
> select FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY [1] (patch 1)
> ==========================================================
> 
> In RISC-V, -fpatchable-function-entry option is used to support
> dynamic ftrace in this commit afc76b8b8011 ("riscv: Using
> PATCHABLE_FUNCTION_ENTRY instead of MCOUNT"). So recordmcount
> don't have to be called to create the __mcount_loc section before
> the vmlinux linking.
> 
> Here selects FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY to tell
> Makefile not to run recordmcount.
> 
> Make function graph use ftrace directly [2] (patch 2)
> ========================================================
> 
> In RISC-V architecture, when we enable the ftrace_graph tracer on some
> functions, the function tracings on other functions will suffer extra
> graph tracing work. In essence, graph_ops isn't limited by its func_hash
> due to the global ftrace_graph_[regs]_call label. That should be
> corrected.
> 
> What inspires me is the commit 0c0593b45c9b ("x86/ftrace: Make function
> graph use ftrace directly") that uses graph_ops::func function to
> install return_hooker and makes the function called against its
> func_hash.
> 
> Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
> ==============================================
> 
> This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS support for RISC-V.
> SAMPLE_FTRACE_DIRECT and SAMPLE_FTRACE_DIRECT_MULTI are also included
> here as the samples for testing DIRECT_CALLS related interface.
> 
> First, select the DYNAMIC_FTRACE_WITH_DIRECT_CALLS to provide
> register_ftrace_direct[_multi] interfaces allowing user to register
> the customed trampoline (direct_caller) as the mcount for one or
> more target functions. And modify_ftrace_direct[_multi] are also
> provided for modify direct_caller.
> 
> At the same time, the samples in ./samples/ftrace/ can be built
> as kerenl module for testing these interfaces with SAMPLE_FTRACE_DIRECT
> and SAMPLE_FTRACE_DIRECT_MULTI selected.
> 
> Second, to make the direct_caller and the other ftrace hooks
> (eg. function/fgraph tracer, k[ret]probes) co-exist, a temporary
> register
> are nominated to store the address of direct_caller in
> ftrace_regs_caller.
> After the setting of the address direct_caller by direct_ops->func and
> the RESTORE_REGS in ftrace_regs_caller, direct_caller will be jumped to
> by the `jr` inst.
> 
> The series's old changes related these patches
> ==========================================
> 
> Changes in v10:
> https://lore.kernel.org/all/20230511093234.3123181-1-suagrfillet@gmail.com/
> 
> - add Acked-by from Björn Töpel in patch 2 and patch 4
> - replace `move` with `mv` in patch3
> - prettify patch 2/4 with proper tabs
> 
> Changes in v9:
> https://lore.kernel.org/linux-riscv/20230510101857.2953955-1-suagrfillet@gmail.com/
> 
> 1. add Acked-by from Björn Töpel in patch 1
> 
> 2. rebase patch2/patch3 on Linux v6.4-rc1
> 
>    - patch 2: to make the `SAVE_ABI_REGS` configurable, revert the
>      modification of mcount-dyn.S from commit (45b32b946a97 "riscv:
> entry: Consolidate general regs saving/restoring")
> 
>    - patch 3: to pass the trace_selftest, add the implement of
>      `ftrace_stub_direct_tramp` from commit (fee86a4ed536 "ftrace:
> selftest: remove broken trace_direct_tramp") ; and fixup the context
> conflict in Kconfig
> 
> Changes in v8:
> https://lore.kernel.org/linux-riscv/20230324033342.3177979-1-suagrfillet@gmail.com/
>   - Fix incorrect address values in the 4nd patch
>   - Rebased on v6.3-rc2
> 
> Changes in v7:
> https://lore.kernel.org/linux-riscv/20230112090603.1295340-1-guoren@kernel.org/
>   - Fixup RESTORE_ABI_REGS by remove PT_T0(sp) overwrite.
>   - Add FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY [1]
>   - Fixup kconfig with HAVE_SAMPLE_FTRACE_DIRECT &
>     HAVE_SAMPLE_FTRACE_DIRECT_MULTI
> 
> Changes in v6:
> https://lore.kernel.org/linux-riscv/20230107133549.4192639-1-guoren@kernel.org/
>   - Replace 8 with MCOUNT_INSN_SIZE
>   - Replace "REG_L a1, PT_RA(sp)" with "mv a1, ra"
>   - Add Evgenii Shatokhin comment
> 
> Changes in v5:
> https://lore.kernel.org/linux-riscv/20221208091244.203407-1-guoren@kernel.org/
>   - Sort Kconfig entries in alphabetical order.
> 
> Changes in v4:
> https://lore.kernel.org/linux-riscv/20221129033230.255947-1-guoren@kernel.org/
>   - Include [3] for maintenance. [Song Shuai]
> 
> Changes in V3:
> https://lore.kernel.org/linux-riscv/20221123153950.2911981-1-guoren@kernel.org/
>   - Include [2] for maintenance. [Song Shuai]
> 
> [1]: https://lore.kernel.org/linux-riscv/CAAYs2=j3Eak9vU6xbAw0zPuoh00rh8v5C2U3fePkokZFibWs2g@mail.gmail.com/T/#t
> [2]: https://lore.kernel.org/lkml/20221120084230.910152-1-suagrfillet@gmail.com/
> [3]: https://lore.kernel.org/linux-riscv/20221123142025.1504030-1-suagrfillet@gmail.com/
> 
> Song Shuai (5):
>    riscv: select FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
>    riscv: ftrace: Add ftrace_graph_func
>    riscv: ftrace: Add DYNAMIC_FTRACE_WITH_DIRECT_CALLS support
>    samples: ftrace: Add riscv support for SAMPLE_FTRACE_DIRECT[_MULTI]
>    samples: ftrace: Make the riscv samples support RV32I
> 
>   arch/riscv/Kconfig                          |   4 +
>   arch/riscv/include/asm/ftrace.h             |  19 +-
>   arch/riscv/kernel/ftrace.c                  |  30 ++-
>   arch/riscv/kernel/mcount-dyn.S              | 200 ++++++++++++++++----
>   samples/ftrace/ftrace-direct-modify.c       |  35 ++++
>   samples/ftrace/ftrace-direct-multi-modify.c |  41 ++++
>   samples/ftrace/ftrace-direct-multi.c        |  25 +++
>   samples/ftrace/ftrace-direct-too.c          |  28 +++
>   samples/ftrace/ftrace-direct.c              |  24 +++
>   9 files changed, 350 insertions(+), 56 deletions(-)
>
Conor Dooley July 6, 2023, 9:53 a.m. UTC | #2
On Thu, Jul 06, 2023 at 05:35:49PM +0800, Song Shuai wrote:
> Ping...

A context-less ping is not very helpful - what are you looking for here?
More reviews? For example, someone to look at 5/5?

> 在 2023/6/27 19:16, Song Shuai 写道:

If it's application you want, you sent the patch only last week - which
was during the merge window, making it unlikely to be applied.

Either way, please try to explain what it is that you are looking for
when you do a ping!

Cheers,
Conor.
Song Shuai July 6, 2023, 10:10 a.m. UTC | #3
在 2023/7/6 17:53, Conor Dooley 写道:
> On Thu, Jul 06, 2023 at 05:35:49PM +0800, Song Shuai wrote:
>> Ping...
> 
> A context-less ping is not very helpful - what are you looking for here?
> More reviews? For example, someone to look at 5/5? >

Sorry for the context-less ping. I hoped someone could look at the 5th 
patch.

>> 在 2023/6/27 19:16, Song Shuai 写道:
> 
> If it's application you want, you sent the patch only last week - which
> was during the merge window, making it unlikely to be applied.
>  > Either way, please try to explain what it is that you are looking for
> when you do a ping!

Thanks for your correction, I'll follow this thread after the merge
window closes.

> 
> Cheers,
> Conor.
>
Björn Töpel July 12, 2023, 6:11 p.m. UTC | #4
Song Shuai <suagrfillet@gmail.com> writes:

[...]

> Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
> ==============================================

We've had some offlist discussions, so here's some input for a wider
audience! Most importantly, this is for Palmer, so that this series is
not merged until a proper BPF trampoline fix is in place.

Note that what's currently usable from BPF trampoline *works*. It's
when this series is added that it breaks.

TL;DR This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS, which enables
fentry/fexit BPF trampoline support. Unfortunately the
fexit/BPF_TRAMP_F_SKIP_FRAME parts of the RV BPF trampoline breaks
with this addition, and need to be addressed *prior* merging this
series. An easy way to reproduce, is just calling any of the kselftest
tests that uses fexit patching.

The issue is around the nop seld, and how a call is done; The nop sled
(patchable-function-entry) size changed from 16B to 8B in commit
6724a76cff85 ("riscv: ftrace: Reduce the detour code size to half"), but
BPF code still uses the old 16B. So it'll work for BPF programs, but not
for regular kernel functions.

An example:

  | ffffffff80fa4150 <bpf_fentry_test1>:
  | ffffffff80fa4150:       0001                    nop
  | ffffffff80fa4152:       0001                    nop
  | ffffffff80fa4154:       0001                    nop
  | ffffffff80fa4156:       0001                    nop
  | ffffffff80fa4158:       1141                    add     sp,sp,-16
  | ffffffff80fa415a:       e422                    sd      s0,8(sp)
  | ffffffff80fa415c:       0800                    add     s0,sp,16
  | ffffffff80fa415e:       6422                    ld      s0,8(sp)
  | ffffffff80fa4160:       2505                    addw    a0,a0,1
  | ffffffff80fa4162:       0141                    add     sp,sp,16
  | ffffffff80fa4164:       8082                    ret

is patched to:

  | ffffffff80fa4150:  f70c0297                     auipc   t0,-150208512
  | ffffffff80fa4154:  eb0282e7                     jalr    t0,t0,-336

The return address to bpf_fentry_test1 is stored in t0 at BPF
trampoline entry. Return to the *parent* is in ra. The trampline has
to deal with this.

For BPF_TRAMP_F_SKIP_FRAME/CALL_ORIG, the BPF trampoline will skip too
many bytes, and not correctly handle parent calls.

Further; The BPF trampoline currently has a different way of patching
the nops for BPF programs, than what ftrace does. That should be changed
to match what ftrace does (auipc/jalr t0).

To summarize:
 * Align BPF nop sled with patchable-function-entry: 8B.
 * Adapt BPF trampoline for 8B nop sleds.
 * Adapt BPF trampoline t0 return, ra parent scheme.
 

Cheers,
Björn
Palmer Dabbelt July 12, 2023, 6:26 p.m. UTC | #5
On Wed, 12 Jul 2023 11:11:08 PDT (-0700), bjorn@kernel.org wrote:
> Song Shuai <suagrfillet@gmail.com> writes:
>
> [...]
>
>> Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
>> ==============================================
>
> We've had some offlist discussions, so here's some input for a wider
> audience! Most importantly, this is for Palmer, so that this series is
> not merged until a proper BPF trampoline fix is in place.
>
> Note that what's currently usable from BPF trampoline *works*. It's
> when this series is added that it breaks.
>
> TL;DR This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS, which enables
> fentry/fexit BPF trampoline support. Unfortunately the
> fexit/BPF_TRAMP_F_SKIP_FRAME parts of the RV BPF trampoline breaks
> with this addition, and need to be addressed *prior* merging this
> series. An easy way to reproduce, is just calling any of the kselftest
> tests that uses fexit patching.
>
> The issue is around the nop seld, and how a call is done; The nop sled
> (patchable-function-entry) size changed from 16B to 8B in commit
> 6724a76cff85 ("riscv: ftrace: Reduce the detour code size to half"), but
> BPF code still uses the old 16B. So it'll work for BPF programs, but not
> for regular kernel functions.
>
> An example:
>
>   | ffffffff80fa4150 <bpf_fentry_test1>:
>   | ffffffff80fa4150:       0001                    nop
>   | ffffffff80fa4152:       0001                    nop
>   | ffffffff80fa4154:       0001                    nop
>   | ffffffff80fa4156:       0001                    nop
>   | ffffffff80fa4158:       1141                    add     sp,sp,-16
>   | ffffffff80fa415a:       e422                    sd      s0,8(sp)
>   | ffffffff80fa415c:       0800                    add     s0,sp,16
>   | ffffffff80fa415e:       6422                    ld      s0,8(sp)
>   | ffffffff80fa4160:       2505                    addw    a0,a0,1
>   | ffffffff80fa4162:       0141                    add     sp,sp,16
>   | ffffffff80fa4164:       8082                    ret
>
> is patched to:
>
>   | ffffffff80fa4150:  f70c0297                     auipc   t0,-150208512
>   | ffffffff80fa4154:  eb0282e7                     jalr    t0,t0,-336
>
> The return address to bpf_fentry_test1 is stored in t0 at BPF
> trampoline entry. Return to the *parent* is in ra. The trampline has
> to deal with this.
>
> For BPF_TRAMP_F_SKIP_FRAME/CALL_ORIG, the BPF trampoline will skip too
> many bytes, and not correctly handle parent calls.
>
> Further; The BPF trampoline currently has a different way of patching
> the nops for BPF programs, than what ftrace does. That should be changed
> to match what ftrace does (auipc/jalr t0).
>
> To summarize:
>  * Align BPF nop sled with patchable-function-entry: 8B.
>  * Adapt BPF trampoline for 8B nop sleds.
>  * Adapt BPF trampoline t0 return, ra parent scheme.

Thanks for digging into this one, I agree we need to sort out the BPF 
breakages before we merge this.  Sounds like there's a rabbit hole here, 
but hopefully we can get it sorted out.

I've dropped this from patchwork and such, as we'll need at least 
another spin.

> Cheers,
> Björn
Pu Lehui July 15, 2023, 9:10 a.m. UTC | #6
On 2023/7/13 2:11, Björn Töpel wrote:
> Song Shuai <suagrfillet@gmail.com> writes:
> 
> [...]
> 
>> Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
>> ==============================================
> 
> We've had some offlist discussions, so here's some input for a wider
> audience! Most importantly, this is for Palmer, so that this series is
> not merged until a proper BPF trampoline fix is in place.
> 
> Note that what's currently usable from BPF trampoline *works*. It's
> when this series is added that it breaks.
> 
> TL;DR This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS, which enables
> fentry/fexit BPF trampoline support. Unfortunately the
> fexit/BPF_TRAMP_F_SKIP_FRAME parts of the RV BPF trampoline breaks
> with this addition, and need to be addressed *prior* merging this
> series. An easy way to reproduce, is just calling any of the kselftest
> tests that uses fexit patching.
> 
> The issue is around the nop seld, and how a call is done; The nop sled
> (patchable-function-entry) size changed from 16B to 8B in commit
> 6724a76cff85 ("riscv: ftrace: Reduce the detour code size to half"), but
> BPF code still uses the old 16B. So it'll work for BPF programs, but not
> for regular kernel functions.
> 
> An example:
> 
>    | ffffffff80fa4150 <bpf_fentry_test1>:
>    | ffffffff80fa4150:       0001                    nop
>    | ffffffff80fa4152:       0001                    nop
>    | ffffffff80fa4154:       0001                    nop
>    | ffffffff80fa4156:       0001                    nop
>    | ffffffff80fa4158:       1141                    add     sp,sp,-16
>    | ffffffff80fa415a:       e422                    sd      s0,8(sp)
>    | ffffffff80fa415c:       0800                    add     s0,sp,16
>    | ffffffff80fa415e:       6422                    ld      s0,8(sp)
>    | ffffffff80fa4160:       2505                    addw    a0,a0,1
>    | ffffffff80fa4162:       0141                    add     sp,sp,16
>    | ffffffff80fa4164:       8082                    ret
> 
> is patched to:
> 
>    | ffffffff80fa4150:  f70c0297                     auipc   t0,-150208512
>    | ffffffff80fa4154:  eb0282e7                     jalr    t0,t0,-336
> 
> The return address to bpf_fentry_test1 is stored in t0 at BPF
> trampoline entry. Return to the *parent* is in ra. The trampline has
> to deal with this.
> 
> For BPF_TRAMP_F_SKIP_FRAME/CALL_ORIG, the BPF trampoline will skip too
> many bytes, and not correctly handle parent calls.
> 
> Further; The BPF trampoline currently has a different way of patching
> the nops for BPF programs, than what ftrace does. That should be changed
> to match what ftrace does (auipc/jalr t0).
> 
> To summarize:
>   * Align BPF nop sled with patchable-function-entry: 8B.
>   * Adapt BPF trampoline for 8B nop sleds.
>   * Adapt BPF trampoline t0 return, ra parent scheme.
> 

Thanks Björn, I make a adaptation as follows, looking forward to your 
review.

https://lore.kernel.org/bpf/20230715090137.2141358-1-pulehui@huaweicloud.com/

> 
> Cheers,
> Björn
> 
>
Björn Töpel Aug. 23, 2023, 8:20 p.m. UTC | #7
Palmer Dabbelt <palmer@rivosinc.com> writes:

> On Wed, 12 Jul 2023 11:11:08 PDT (-0700), bjorn@kernel.org wrote:
>> Song Shuai <suagrfillet@gmail.com> writes:
>>
>> [...]
>>
>>> Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
>>> ==============================================
>>
>> We've had some offlist discussions, so here's some input for a wider
>> audience! Most importantly, this is for Palmer, so that this series is
>> not merged until a proper BPF trampoline fix is in place.
>>
>> Note that what's currently usable from BPF trampoline *works*. It's
>> when this series is added that it breaks.
>>
>> TL;DR This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS, which enables
>> fentry/fexit BPF trampoline support. Unfortunately the
>> fexit/BPF_TRAMP_F_SKIP_FRAME parts of the RV BPF trampoline breaks
>> with this addition, and need to be addressed *prior* merging this
>> series. An easy way to reproduce, is just calling any of the kselftest
>> tests that uses fexit patching.
>>
>> The issue is around the nop seld, and how a call is done; The nop sled
>> (patchable-function-entry) size changed from 16B to 8B in commit
>> 6724a76cff85 ("riscv: ftrace: Reduce the detour code size to half"), but
>> BPF code still uses the old 16B. So it'll work for BPF programs, but not
>> for regular kernel functions.
>>
>> An example:
>>
>>   | ffffffff80fa4150 <bpf_fentry_test1>:
>>   | ffffffff80fa4150:       0001                    nop
>>   | ffffffff80fa4152:       0001                    nop
>>   | ffffffff80fa4154:       0001                    nop
>>   | ffffffff80fa4156:       0001                    nop
>>   | ffffffff80fa4158:       1141                    add     sp,sp,-16
>>   | ffffffff80fa415a:       e422                    sd      s0,8(sp)
>>   | ffffffff80fa415c:       0800                    add     s0,sp,16
>>   | ffffffff80fa415e:       6422                    ld      s0,8(sp)
>>   | ffffffff80fa4160:       2505                    addw    a0,a0,1
>>   | ffffffff80fa4162:       0141                    add     sp,sp,16
>>   | ffffffff80fa4164:       8082                    ret
>>
>> is patched to:
>>
>>   | ffffffff80fa4150:  f70c0297                     auipc   t0,-150208512
>>   | ffffffff80fa4154:  eb0282e7                     jalr    t0,t0,-336
>>
>> The return address to bpf_fentry_test1 is stored in t0 at BPF
>> trampoline entry. Return to the *parent* is in ra. The trampline has
>> to deal with this.
>>
>> For BPF_TRAMP_F_SKIP_FRAME/CALL_ORIG, the BPF trampoline will skip too
>> many bytes, and not correctly handle parent calls.
>>
>> Further; The BPF trampoline currently has a different way of patching
>> the nops for BPF programs, than what ftrace does. That should be changed
>> to match what ftrace does (auipc/jalr t0).
>>
>> To summarize:
>>  * Align BPF nop sled with patchable-function-entry: 8B.
>>  * Adapt BPF trampoline for 8B nop sleds.
>>  * Adapt BPF trampoline t0 return, ra parent scheme.
>
> Thanks for digging into this one, I agree we need to sort out the BPF 
> breakages before we merge this.  Sounds like there's a rabbit hole here, 
> but hopefully we can get it sorted out.
>
> I've dropped this from patchwork and such, as we'll need at least 
> another spin.

Palmer,

The needed BPF patch is upstream in the bpf-next tree, and has been for
a couple of weeks.

I think this series is a candidate for RISC-V -next! It would help
RISC-V BPF a lot in terms of completeness.


Björn
Björn Töpel Aug. 30, 2023, 3:28 p.m. UTC | #8
Björn Töpel <bjorn@kernel.org> writes:

> Palmer Dabbelt <palmer@rivosinc.com> writes:
>
>> On Wed, 12 Jul 2023 11:11:08 PDT (-0700), bjorn@kernel.org wrote:
>>> Song Shuai <suagrfillet@gmail.com> writes:
>>>
>>> [...]
>>>
>>>> Add WITH_DIRECT_CALLS support [3] (patch 3, 4)
>>>> ==============================================
>>>
>>> We've had some offlist discussions, so here's some input for a wider
>>> audience! Most importantly, this is for Palmer, so that this series is
>>> not merged until a proper BPF trampoline fix is in place.
>>>
>>> Note that what's currently usable from BPF trampoline *works*. It's
>>> when this series is added that it breaks.
>>>
>>> TL;DR This series adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS, which enables
>>> fentry/fexit BPF trampoline support. Unfortunately the
>>> fexit/BPF_TRAMP_F_SKIP_FRAME parts of the RV BPF trampoline breaks
>>> with this addition, and need to be addressed *prior* merging this
>>> series. An easy way to reproduce, is just calling any of the kselftest
>>> tests that uses fexit patching.
>>>
>>> The issue is around the nop seld, and how a call is done; The nop sled
>>> (patchable-function-entry) size changed from 16B to 8B in commit
>>> 6724a76cff85 ("riscv: ftrace: Reduce the detour code size to half"), but
>>> BPF code still uses the old 16B. So it'll work for BPF programs, but not
>>> for regular kernel functions.
>>>
>>> An example:
>>>
>>>   | ffffffff80fa4150 <bpf_fentry_test1>:
>>>   | ffffffff80fa4150:       0001                    nop
>>>   | ffffffff80fa4152:       0001                    nop
>>>   | ffffffff80fa4154:       0001                    nop
>>>   | ffffffff80fa4156:       0001                    nop
>>>   | ffffffff80fa4158:       1141                    add     sp,sp,-16
>>>   | ffffffff80fa415a:       e422                    sd      s0,8(sp)
>>>   | ffffffff80fa415c:       0800                    add     s0,sp,16
>>>   | ffffffff80fa415e:       6422                    ld      s0,8(sp)
>>>   | ffffffff80fa4160:       2505                    addw    a0,a0,1
>>>   | ffffffff80fa4162:       0141                    add     sp,sp,16
>>>   | ffffffff80fa4164:       8082                    ret
>>>
>>> is patched to:
>>>
>>>   | ffffffff80fa4150:  f70c0297                     auipc   t0,-150208512
>>>   | ffffffff80fa4154:  eb0282e7                     jalr    t0,t0,-336
>>>
>>> The return address to bpf_fentry_test1 is stored in t0 at BPF
>>> trampoline entry. Return to the *parent* is in ra. The trampline has
>>> to deal with this.
>>>
>>> For BPF_TRAMP_F_SKIP_FRAME/CALL_ORIG, the BPF trampoline will skip too
>>> many bytes, and not correctly handle parent calls.
>>>
>>> Further; The BPF trampoline currently has a different way of patching
>>> the nops for BPF programs, than what ftrace does. That should be changed
>>> to match what ftrace does (auipc/jalr t0).
>>>
>>> To summarize:
>>>  * Align BPF nop sled with patchable-function-entry: 8B.
>>>  * Adapt BPF trampoline for 8B nop sleds.
>>>  * Adapt BPF trampoline t0 return, ra parent scheme.
>>
>> Thanks for digging into this one, I agree we need to sort out the BPF 
>> breakages before we merge this.  Sounds like there's a rabbit hole here, 
>> but hopefully we can get it sorted out.
>>
>> I've dropped this from patchwork and such, as we'll need at least 
>> another spin.
>
> Palmer,
>
> The needed BPF patch is upstream in the bpf-next tree, and has been for
> a couple of weeks.
>
> I think this series is a candidate for RISC-V -next! It would help
> RISC-V BPF a lot in terms of completeness.

Palmer,

The needed fix for BPF is now in Linus' tree, commit 25ad10658dc1
("riscv, bpf: Adapt bpf trampoline to optimized riscv ftrace
framework"). IOW, this ftrace series can be merged now.


Björn