- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
After a great deal of work along with help from Altera and the Card vendor, we now have a working kernel flashed onto the card. We were also able to compile the host program and as such we attempted to run it. When we did we got the following error messages.
First:WARNING: NOT using DMA for transfer of size 1024 because of lack of alignment src ptr = 0x7fa01dffd850, dst ptr = 0x2010080, reading = 0
000000000000000000000000000000000000000000000000e4ae120000000000
followed by: WARNING: NOT using DMA for transfer of size 1024 because of lack of alignment
src ptr = 0x2010080, dst ptr = 0x7fa01dffd850, reading = 1
neither of these message are coming from the host program. Does anyone have any insight into what these messages mean? -Aaron
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
The address alignment problem ("WARNING: NOT using DMA for transfer of size 1024 because of lack of alignment"). When we malloc memory for the host, please align the address first before use in the host program.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
--- Quote Start --- The address alignment problem ("WARNING: NOT using DMA for transfer of size 1024 because of lack of alignment"). When we malloc memory for the host, please align the address first before use in the host program. --- Quote End --- Exactly. It's explained in the optimization guide (http://www.altera.com/literature/hb/opencl-sdk/aocl_optimization_guide.pdf), page 1-7. It says the data should be 64-byte aligned. In my case, without proper alignement, one of my memory transfers and kernel execution takes ~300secs. With alignment, it gets down to ~3sec. So you better align it. :D
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
If you take a look at any of the example designs (except hello world) you'll see code fragments that you can use to align these buffers on Windows and Linux hosts. I recommend making the alignment a macro just in case you change to a platform that requires an even larger alignment.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Also please see our optimization guide (http://www.altera.com/literature/hb/opencl-sdk/aocl_optimization_guide.pdf), section "Aligned Memory Allocation".
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page