Hi, I'm implementing a 2D graphics driver for an ARM A9 multicore SoC. As part of this, I'm using remap_pfn_range() to map memory to user space. During testing I realized that remap_pfn_range() flushes the L1 cache, but doesn't flush the L2 cache. As a result, I can get L2 cache hits even if I mapped the memory uncached.
What is the motivation for this? Any help on this is greatly appreciated. Best regards, Oliver P.S. I'm sorry if this is a bit of topic for this mailing list, but I didn't know where else to post this.
