Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   David Laight <David.Laight@ACULAB.COM>
To:     'Jisheng Zhang' <jszhang@kernel.org>,
        Paul Walmsley <paul.walmsley@sifive.com>,
        Palmer Dabbelt <palmer@dabbelt.com>,
        Albert Ou <aou@eecs.berkeley.edu>
CC:     "linux-riscv@lists.infradead.org" <linux-riscv@lists.infradead.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: RE: [RFC PATCH] riscv: reduce THREAD_SIZE from 16KB to 8KB for RV64
Thread-Topic: [RFC PATCH] riscv: reduce THREAD_SIZE from 16KB to 8KB for RV64
Thread-Index: AQHYG4I1cWK2Agm31k+W40NUWVDrs6yHbd8Q
Date:   Mon, 7 Feb 2022 03:35:55 +0000
Message-ID: <04326e1a50fa46eaa08df3a5df6f14f6@AcuMS.aculab.com>
References: <20220206174359.2986-1-jszhang@kernel.org>
In-Reply-To: <20220206174359.2986-1-jszhang@kernel.org>
Accept-Language: en-GB, en-US
MIME-Version: 1.0
Content-Language: en-US
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8BIT
Precedence: bulk

From: Jisheng Zhang
> Sent: 06 February 2022 17:44
> 
> After irq stack is supported, it's possible to use small THREAD_SIZE.
> In fact, I tested this patch on a Lichee RV board, looks good so far.

Is riscv using vmalloc with a guard page?

You won't find the deepest kernel stack use with trivial tests.
I'd hazard a guess that it is inside printk() in some error path.
Debugging stack overflow is horrid.

With no alloca() no recursion and limited indirect calls it is
technically possible to statically calculate maximum stack use.
But I don't think anyone has tried to do it for the linux kernel.
I did do it for an embedded system (that had almost no indirect calls)
and found we didn't have enough memory for the required stacks!

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)