On 29/11/2012 06:28, Wolfgang Denk wrote: > Dear Troy Kisky, > > In message <50b6cb79.4030...@boundarydevices.com> you wrote: >> >> Would you like to see the Linux way of ALIGN, or ROUND? > > Do you align some buffer or similar, or do you round (up) a size? > >> Now, back to the other topic you raised. Should I apply the bug work-around >> for all version 2 headers, or find a way to distinguish mx53/mx6?
> > I cannot tell. Stefano, what do you think? Well, I am thinking about which are the real benefits. If we always round up the size to 512 bytes for V2 header, *maybe* we constrain the i.MX6 to load some bytes more, but it is the only drawback. And this if the i.MX6 does not suffer of the same problem found on i.MX53. On the other side, having two different versions of V2 header is confusing. It is then undocumented by Freescale, and maybe it is possible to find a note in some errata. Having the same interface without special hacking for each SOC overcomes the increment in the footprint for the i.MX6 (in worst case 511 bytes - and not a lot compared to current size of U-Boot for MX5/MX6, usually several hundred of KB). Best regards, Stefano Babic -- ===================================================================== DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sba...@denx.de ===================================================================== _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
