[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <7adc9e19-7ffc-4b11-3e18-6e3a5225638f@redhat.com>
Date: Tue, 21 Jun 2022 11:08:21 -0400
From: Waiman Long <longman@...hat.com>
To: guoren@...nel.org, palmer@...osinc.com, arnd@...db.de,
peterz@...radead.org, boqun.feng@...il.com,
Conor.Dooley@...rochip.com, chenhuacai@...ngson.cn,
kernel@...0n.name, r@....cc, shorne@...il.com
Cc: linux-riscv@...ts.infradead.org, linux-arch@...r.kernel.org,
linux-kernel@...r.kernel.org, Guo Ren <guoren@...ux.alibaba.com>
Subject: Re: [PATCH V6 0/2] riscv: Support qspinlock with generic headers
On 6/21/22 10:49, guoren@...nel.org wrote:
> From: Guo Ren <guoren@...ux.alibaba.com>
>
> Enable qspinlock and meet the requirements mentioned in a8ad07e5240c9
> ("asm-generic: qspinlock: Indicate the use of mixed-size atomics").
>
> RISC-V LR/SC pairs could provide a strong/weak forward guarantee that
> depends on micro-architecture. And RISC-V ISA spec has given out
> several limitations to let hardware support strict forward guarantee
> (RISC-V User ISA - 8.3 Eventual Success of Store-Conditional
> Instructions):
> We restricted the length of LR/SC loops to fit within 64 contiguous
> instruction bytes in the base ISA to avoid undue restrictions on
> instruction cache and TLB size and associativity. Similarly, we
Does the 64 contiguous bytes need to be cacheline aligned?
Regards,
Longman
Powered by blists - more mailing lists