[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20190604120623epcms5p3fc13e98047e7b44d6da144425213b4fe@epcms5p3>
Date: Tue, 04 Jun 2019 17:36:23 +0530
From: Vaneet Narang <v.narang@...sung.com>
To: Andrew Morton <akpm@...ux-foundation.org>,
Maninder Singh <maninder1.s@...sung.com>
CC: "herbert@...dor.apana.org.au" <herbert@...dor.apana.org.au>,
"davem@...emloft.net" <davem@...emloft.net>,
"keescook@...omium.org" <keescook@...omium.org>,
"gustavo@...eddedor.com" <gustavo@...eddedor.com>,
"joe@...ches.com" <joe@...ches.com>,
"linux-crypto@...r.kernel.org" <linux-crypto@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
AMIT SAHRAWAT <a.sahrawat@...sung.com>,
PANKAJ MISHRA <pankaj.m@...sung.com>,
Vaneet Narang <v.narang@...sung.com>
Subject: RE:(2) [PATCH 0/4] zstd: reduce stack usage
Hi Andrew,
>> This patch set reduces stack usage for zstd code, because target like ARM has
>> limited 8KB kernel stack, which is getting overflowed due to hight stack usage
>> of zstd code with call flow like:
>That's rather bad behaviour. I assume the patchset actually fixes this?
Yes, patchset tries to reduce around 300 bytes of stack usage of zstd compression path.
We faced high stack usage issue on switching compression algo from LZO/LZ4 to zstd algo.
zstd compression uses around 1200 bytes of stack which is huge as compared
to LZO/LZ4 which uses negligible stack (< 200 bytes).
>I think I'll schedule the patchset for 5.2-rcX so that zstd is actually
>usable on arm in 5.2. Does that sound OK?
OK
Regards,
Vaneet Narang
Powered by blists - more mailing lists