ProgrammingThis forum is for all programming questions.
The question does not have to be directly related to Linux and any language is fair game.
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
I've been working with a parallel IO board that sets up its own memory mapped region. I have no trouble reading and processing the data from this buffer. I'm trying to copy this data to the hard drive (large file accesses) and the process is EXTREMELY speed critical. I'm using a UDMA hard drive and the data transfers from the IO card to its memory mapped region use Demand Mode DMA accesses.
Does anyone have any suggestions on what the fastest way to copy these chunks of data is? I've tried memory mapping the file to kernel memory, but it seems like it takes longer on account of remapping the file for every subsequent chunk. (I can only get it to work if it's copied to an intermediate buffer) Here's what seems to work the best so far. It seems too simple, but isn't fast enough. Any suggestions?
/* Telling the parallel IO card (_fd) to copy the data from its buffers to a memory mapped read transfer buffer */
read(_fd, NULL, transferSize);
*** Then... ***
/* Writing the data to disk (fp) from the read transfer buffer (mmRead) */
write(fp, mmRead, transferSize);
I thought about possibly seeing if the IO card would copy directly to a memory mapped file, but I get lock ups and segmentation faults:
read(_fd, mmRead, transferSize);
*** OR ***
memcpy(mmFile, mmRead, transferSize);