Date: Thu, 26 Sep 2002 22:41:31 +0200 From: Mark Santcroos <marks@ripe.net> To: Josef Karthauser <joe@FreeBSD.ORG> Cc: freebsd-hackers@FreeBSD.ORG Subject: Re: vmware reads disk on non-sector boundary Message-ID: <20020926204131.GA1079@laptop.6bone.nl> In-Reply-To: <20020926185036.GB10421@genius.tao.org.uk> References: <20020925173453.GA1347@laptop.6bone.nl> <Pine.BSF.4.21.0209251114130.46717-100000@InterJet.elischer.org> <20020925203930.GB25571@genius.tao.org.uk> <20020926084733.GA26352@spc.org> <20020926085237.GA2645@genius.tao.org.uk> <20020926185036.GB10421@genius.tao.org.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Sep 26, 2002 at 07:50:36PM +0100, Josef Karthauser wrote:
> It took a while to find, but this is the hack I was referring to:
>
> Take a look at /sys/compat/linux/linux_stats.c,
Thanks for taking the effort for looking this up.
However, the function in question - newstat_copyout - is not in 'action'
here. (had set a breakpoint in ddb)
It's still unclear to me what recent change broke vmware.
The removal of block devices Julian referred to is not something recent,
is it?
I mentioned earlier to you that I am writing a linux_read system call.
This will basicly work, as long as I will be able to do the following (in
pseudo code):
linux_read(readsize) {
if (read_is_on_raw_device)
readlen = (readlen % BLOCKSIZE) + BLOCKSIZE;
/* or something like this plus some other mangling of the buffer
*/
read(readlen);
}
The only thing I have to find out is the check to see whether it is a read
on a raw device or not. I strongly prefer to do the check in linux_read
and not in the disk layer code. (It would be easier, but not cleaner)
Mark
--
Mark Santcroos RIPE Network Coordination Centre
http://www.ripe.net/home/mark/ New Projects Group/TTM
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020926204131.GA1079>