linux-kernel - Re: [PATCH v2 6/6] RISC-V: hwprobe: Document unaligned vector perf key

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Fri, 21 Jun 2024 14:30:39 -0400
From: Jesse Taube <jesse@...osinc.com>
To: Evan Green <evan@...osinc.com>
Cc: linux-riscv@...ts.infradead.org, Jonathan Corbet <corbet@....net>, 
	Paul Walmsley <paul.walmsley@...ive.com>, Palmer Dabbelt <palmer@...belt.com>, 
	Albert Ou <aou@...s.berkeley.edu>, Conor Dooley <conor@...nel.org>, Rob Herring <robh@...nel.org>, 
	Krzysztof Kozlowski <krzk+dt@...nel.org>, Clément Léger <cleger@...osinc.com>, 
	Andrew Jones <ajones@...tanamicro.com>, Charlie Jenkins <charlie@...osinc.com>, 
	Xiao Wang <xiao.w.wang@...el.com>, Andy Chiu <andy.chiu@...ive.com>, 
	Eric Biggers <ebiggers@...gle.com>, Greentime Hu <greentime.hu@...ive.com>, 
	Björn Töpel <bjorn@...osinc.com>, 
	Heiko Stuebner <heiko@...ech.de>, Costa Shulyupin <costa.shul@...hat.com>, 
	Andrew Morton <akpm@...ux-foundation.org>, Baoquan He <bhe@...hat.com>, 
	Anup Patel <apatel@...tanamicro.com>, Zong Li <zong.li@...ive.com>, 
	Sami Tolvanen <samitolvanen@...gle.com>, Ben Dooks <ben.dooks@...ethink.co.uk>, 
	Alexandre Ghiti <alexghiti@...osinc.com>, "Gustavo A. R. Silva" <gustavoars@...nel.org>, 
	Erick Archer <erick.archer@....com>, Joel Granados <j.granados@...sung.com>, linux-doc@...r.kernel.org, 
	linux-kernel@...r.kernel.org, devicetree@...r.kernel.org
Subject: Re: [PATCH v2 6/6] RISC-V: hwprobe: Document unaligned vector perf key

On Thu, Jun 20, 2024 at 2:52 PM Evan Green <evan@...osinc.com> wrote:
>
> On Thu, Jun 13, 2024 at 12:18 PM Jesse Taube <jesse@...osinc.com> wrote:
> >
> > Document key for reporting the speed of unaligned vector accesses.
> > The descriptions are the same as the scalar equivalent values.
> >
> > Signed-off-by: Jesse Taube <jesse@...osinc.com>
> > ---
> > V1 -> V2:
> >   - New patch
> > ---
> >  Documentation/arch/riscv/hwprobe.rst | 16 ++++++++++++++++
> >  1 file changed, 16 insertions(+)
> >
> > diff --git a/Documentation/arch/riscv/hwprobe.rst b/Documentation/arch/riscv/hwprobe.rst
> > index 7085a694b801..344bea1e21bd 100644
> > --- a/Documentation/arch/riscv/hwprobe.rst
> > +++ b/Documentation/arch/riscv/hwprobe.rst
> > @@ -236,3 +236,19 @@ The following keys are defined:
> >
> >  * :c:macro:`RISCV_HWPROBE_KEY_ZICBOZ_BLOCK_SIZE`: An unsigned int which
> >    represents the size of the Zicboz block in bytes.
> > +
> > +* :c:macro:`RISCV_HWPROBE_KEY_VEC_MISALIGNED_PERF`: An enum value describing the
> > +  performance of misaligned vector accesses on the selected set of processors.
> > +
> > +  * :c:macro:`RISCV_HWPROBE_VEC_MISALIGNED_UNKNOWN`: The performance of misaligned
> > +    accesses is unknown.
> > +
> > +  * :c:macro:`RISCV_HWPROBE_VEC_MISALIGNED_SLOW`: Misaligned accesses are slower
>
> Should we specify what size of vector access we're comparing against?
> In other words, crispen up what "misaligned access" exactly means. I
> realize you copied this from my text. I really should have said
> "misaligned native word size accesses".

In arch/riscv/kernel/vec-copy-unaligned.S I set WORD_EEW to 32bits.
The rationale for using 32bits is
("riscv: vector: adjust minimum Vector requirement to ZVE32X") in this set.
https://lore.kernel.org/all/20240412-zve-detection-v4-0-e0c45bb6b253@sifive.com/

I'll change faste and slow to start with "32bit misaligned accesses are"

Thanks,
Jesse
>
> > +    than equivalent byte accesses.  Misaligned accesses may be supported
> > +    directly in hardware, or trapped and emulated by software.
> > +
> > +  * :c:macro:`RISCV_HWPROBE_VEC_MISALIGNED_FAST`: Misaligned accesses are faster
> > +    than equivalent byte accesses.
> > +
> > +  * :c:macro:`RISCV_HWPROBE_VEC_MISALIGNED_UNSUPPORTED`: Misaligned accesses are
> > +    not supported at all and will generate a misaligned address fault.
> > --
> > 2.43.0
> >