On 15 April 2015 at 11:55, Matt Fleming <matt@xxxxxxxxxxxxxxxxxxx> wrote: > On Fri, 10 Apr, at 07:14:34PM, Ard Biesheuvel wrote: >> >> It is not such a big deal: the memory is reclaimed anyway, I was just >> trying to reduce the fragmentation a bit, and trying to avoid >> efi_xxx_alloc() which are substantially heavier than calling >> allocate_pool() or allocate_pages() directly. > > Out of curiosity, do you have any runtime numbers to backup the claim > that allocate_*() is substantially more lightweight? > No, I do not. But since efi_[high|low]_alloc() call efi_get_memory_map() internally, which itself calls allocate_pool() at least twice [typically], and then iterate over the entire memory map to select a suitable slot which gets allocated using allocate_pages(), calling either of allocate_[pool|pages] directly is arguably more lightweight. But claiming they are 'substantially heavier' is unsubstantiated. -- Ard. -- To unsubscribe from this list: send the line "unsubscribe linux-efi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
