[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CANXhq0rbKJFHgVCQzXDAmSOi6KjCcwE583tAVF7TNLTpLsiTSQ@mail.gmail.com>
Date: Thu, 27 Aug 2020 16:22:43 +0800
From: Zong Li <zong.li@...ive.com>
To: Palmer Dabbelt <palmer@...belt.com>
Cc: Paul Walmsley <paul.walmsley@...ive.com>,
linux-riscv <linux-riscv@...ts.infradead.org>,
"linux-kernel@...r.kernel.org List" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 0/3] Get cache information from userland
On Fri, Aug 21, 2020 at 4:45 AM Palmer Dabbelt <palmer@...belt.com> wrote:
>
> On Fri, 03 Jul 2020 01:57:52 PDT (-0700), zong.li@...ive.com wrote:
> > There are no standard CSR registers to provide cache information, the
> > way for RISC-V is to get this information from DT. Currently, AT_L1I_X,
> > AT_L1D_X and AT_L2_X are present in glibc header, and sysconf syscall
> > could use them to get information of cache through AUX vector. We
> > exploit 'struct cacheinfo' to obtain the information of cache, then we
> > don't need additional variable or data structure to record it.
> >
> > We also need some works in glibc, but we have to support the function in
> > kernel first by rule of glibc, then post the patch to glibc site.
> >
> > The result of 'getconf -a' as follows:
> >
> > LEVEL1_ICACHE_SIZE 32768
> > LEVEL1_ICACHE_ASSOC 8
> > LEVEL1_ICACHE_LINESIZE 64
> > LEVEL1_DCACHE_SIZE 32768
> > LEVEL1_DCACHE_ASSOC 8
> > LEVEL1_DCACHE_LINESIZE 64
> > LEVEL2_CACHE_SIZE 2097152
> > LEVEL2_CACHE_ASSOC 32
> > LEVEL2_CACHE_LINESIZE 64
> >
> > Zong Li (3):
> > riscv: Set more data to cacheinfo
> > riscv: Define AT_VECTOR_SIZE_ARCH for ARCH_DLINFO
> > riscv: Add cache information in AUX vector
> >
> > arch/riscv/include/asm/cacheinfo.h | 14 +++++
> > arch/riscv/include/asm/elf.h | 13 ++++
> > arch/riscv/include/uapi/asm/auxvec.h | 24 ++++++++
> > arch/riscv/kernel/cacheinfo.c | 92 +++++++++++++++++++++++-----
> > 4 files changed, 127 insertions(+), 16 deletions(-)
> > create mode 100644 arch/riscv/include/asm/cacheinfo.h
>
> Sorry, I lost track of these. There's a little issue in #1, LMK if you want to
> send another version or if you want me to just fix it.
Sorry for being late here, I have sent the second version patches to
fix it, thanks for your reviewing.
Powered by blists - more mailing lists