Received-SPF: pass (google.com: domain of linux-kernel+bounces-117059-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Precedence: bulk
MIME-Version: 1.0
References: <20240306165904.108141-1-puranjay12@gmail.com> <87ttlhdeqb.fsf@all.your.base.are.belong.to.us>
 <ZfBbxPDd0rz6FN2T@FVFF77S0Q05N> <8734suqsth.fsf@all.your.base.are.belong.to.us>
 <mb61pplvw6grf.fsf@gmail.com> <87zfv0onre.fsf@all.your.base.are.belong.to.us>
 <87il1oedx8.fsf@all.your.base.are.belong.to.us> <CABgGipWPuvwi43v1+60-=0_MN_q_CD0ZGasxHHVWJ37cig5MmA@mail.gmail.com>
 <87msqsotr8.fsf@all.your.base.are.belong.to.us> <CABgGipUayONdvaH0nexBJUai=qT3yqURhinTbDWxORfMDQ_5pw@mail.gmail.com>
In-Reply-To: <CABgGipUayONdvaH0nexBJUai=qT3yqURhinTbDWxORfMDQ_5pw@mail.gmail.com>
From: Robbin Ehn <rehn@rivosinc.com>
Date: Mon, 25 Mar 2024 13:50:11 +0100
Message-ID: <CAEEvPLB0eVXwRxucKDtc4mG11BwueV4pk4g9Zi82MPeTaP41uw@mail.gmail.com>
Subject: Re: [RFC PATCH] riscv: Implement HAVE_DYNAMIC_FTRACE_WITH_CALL_OPS
To: Andy Chiu <andy.chiu@sifive.com>
Cc: =?UTF-8?B?QmrDtnJuIFTDtnBlbA==?= <bjorn@kernel.org>, 
	Puranjay Mohan <puranjay12@gmail.com>, Mark Rutland <mark.rutland@arm.com>, 
	Paul Walmsley <paul.walmsley@sifive.com>, Palmer Dabbelt <palmer@dabbelt.com>, 
	Albert Ou <aou@eecs.berkeley.edu>, Steven Rostedt <rostedt@goodmis.org>, 
	Masami Hiramatsu <mhiramat@kernel.org>, Sami Tolvanen <samitolvanen@google.com>, 
	Guo Ren <guoren@kernel.org>, Ley Foon Tan <leyfoon.tan@starfivetech.com>, 
	Deepak Gupta <debug@rivosinc.com>, Sia Jee Heng <jeeheng.sia@starfivetech.com>, 
	Bjorn Topel <bjorn@rivosinc.com>, Song Shuai <suagrfillet@gmail.com>, 
	"Cl'ement L'eger" <cleger@rivosinc.com>, Al Viro <viro@zeniv.linux.org.uk>, 
	Jisheng Zhang <jszhang@kernel.org>, linux-riscv@lists.infradead.org, 
	linux-kernel@vger.kernel.org, linux-trace-kernel@vger.kernel.org, 
	Brendan Sweeney <brs@rivosinc.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hey,

> <index-in-dispatch-list>
> func_symbol:
> auipc t0, common_dispatch:high <=3D> j actual_func:
> jalr t0, common_dispatch:low(t0)
>

If you are patching in a jump, I don't see why you wouldn't jump over
ld+jalr? (no need for common dispatch)
Patching jalr with nop, and keeping auipc, addresses the issue with
having to jump in the disabled case.
But needs either common dispatch or per func dispatch.

Thanks, Robbin

> common_dispatch:
> load t1, index + dispatch-list
> ld t1, 0(t1)
> jr t1
>
>
> >
> > > However, one thing I am not very sure is: do we need a destination
> > > address in a "per-function" manner? It seems like most of the time th=
e
> > > destination address can only be ftrace_call, or ftrace_regs_call. If
> > > the number of destination addresses is very few, then we could
> > > potentially reduce the size of
> > > <func_trace_target_data_8B_per_function>.
> >
> > Yes, we do need a per-function manner. BPF, e.g., uses
> > dynamically/JIT:ed trampolines/targets.
> >
> >
> >
> > Bj=C3=B6rn
>
> Cheers,
> Andy