Date: Thu, 18 May 2000 11:45:06 -0400 From: Joy Ganguly <joy@niksun.com> To: Matthew Dillon <dillon@apollo.backplane.com>, freebsd-hackers <freebsd-hackers@freeBSD.org> Subject: Re: Question about file write???? Message-ID: <39241002.1F03763A@falcon.niksun.com> References: <20000517160746.12768.qmail@www0a.netaddress.usa.net> <20000517102516.B14968@fw.wintelcom.net> <3922DB0F.9D9095C9@falcon.niksun.com> <20000517135644.D14968@fw.wintelcom.net> <39231E06.31DFB153@falcon.niksun.com> <200005172324.QAA28046@apollo.backplane.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Dillon wrote:
> I think it would be a whole lot easier to implement mmap() on the
> device, then allow the user process to mmap() the device buffer
> and read() and write() it directly. I don't think you need to
> implement any ioctl's for this problem.
>
> fd = open device
> ptr = mmap(...)
>
> fd = open file
> read(fd, ptr, BUFSIZE) read from file, store into device buf
> ...
> write(fd, ptr, BUFSIZE) read from device buf, store into file
>
>
I can do that but will it improve performance? right now the ioctl
which does the write takes substantially more time than the normal write from
user space. Is this because the previous read touches the user space buffer and
hence it is in the cache?? i am not very sure on this . any suggestions??
thanx in advance
joy
>
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?39241002.1F03763A>