By the way on MacOSX we also pre-allocate some buffer in userspace first for everything. Well it's up to you anyway I haven't seen a move into a good direction in that area for years anyway on Linux so I'm used to deal with what is currently available. On Tue, Nov 17, 2015 at 8:13 PM, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote: > On Tue, 17 Nov 2015, Steinar H. Gunderson wrote: > >> On Tue, Nov 17, 2015 at 07:17:31PM +0100, Markus Rechberger wrote: >> > 1. the memset on isochronous transfers to empty the buffers in order >> > to avoid leaking raw memory to userspace (this costs a lot on intel >> > Atoms and is also noticeable on other systems). >> > >> > 2. the memory fragmentation. Seems like recent systems have a better >> > performance here since we did not get that report for several months >> > now, or maybe the user behavior changed. >> > Some older Linux systems (maybe 2-3 years old) triggered this issue >> > way more often. >> >> I guess if we get transparent zerocopy, both of these are going away >> just like with your patch, right? The only difference is really who sets up >> the memory area (the kernel or not). >> >> Alan, could we perhaps let the zerocopy flag make the request fail (instead >> of going through a bounce buffer) if direct DMA is not possible? That way, >> it would be quite obvious that you need to allocate the memory some other way >> instead of silently hitting the issues Markus mention. > > But what other way of allocating memory is there? > > With scatter-gather lists, fragmentation isn't an issue. But bounce > buffers are unavoidable if the memory isn't accessible to the USB > hardware. > > Alan Stern > -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
