[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170215203734.GC5531@obsidianresearch.com>
Date: Wed, 15 Feb 2017 13:37:34 -0700
From: Jason Gunthorpe <jgunthorpe@...idianresearch.com>
To: matthew.gerlach@...ux.intel.com
Cc: Alan Tull <delicious.quinoa@...il.com>,
Moritz Fischer <moritz.fischer@...us.com>,
linux-kernel <linux-kernel@...r.kernel.org>,
linux-fpga@...r.kernel.org
Subject: Re: [RFC 7/8] fpga-region: add sysfs interface
On Wed, Feb 15, 2017 at 12:07:15PM -0800, matthew.gerlach@...ux.intel.com wrote:
> The format of the meta data associated with a fpga bitstream is certainly a
> subject on its own. HTTP style plain text is definately easy to understand
> and more importantly it is extendable. On the other hand, it seems
> dangerous to be doing a lot of string parsing in the kernel.
It is fairly close to binary parsing.. The process is
- Find the first occurance of \n\n, must be less than XX bytes
- Memcpy that from the sg list into a linear buffer
- Replace all \n with \0
To access a key:
- Case insensitive search for START + "Key: " or \0 + "Key: "
- Return as a string the part after the match
This isn't the sort of string parsing that typically gets you into
trouble. If we can't code the above correctly then we will screw up
safe binary parsing of strings too :)
Jason
Powered by blists - more mailing lists