[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-devel] [PATCH 09/12] arm: load dom0 kernel from first boot module
On Thu, 2012-11-29 at 17:15 +0000, Tim Deegan wrote: > At 16:23 +0000 on 13 Nov (1352823801), Ian Campbell wrote: > > -static int kernel_try_zimage_prepare(struct kernel_info *info) > > +static int kernel_try_zimage_prepare(struct kernel_info *info, > > + paddr_t addr, paddr_t size) > > { > > uint32_t *zimage = (void *)FIXMAP_ADDR(FIXMAP_MISC); > > uint32_t start, end; > > struct minimal_dtb_header dtb_hdr; > > > > - set_fixmap(FIXMAP_MISC, KERNEL_FLASH_ADDRESS >> PAGE_SHIFT, > > DEV_SHARED); > > + set_fixmap(FIXMAP_MISC, addr >> PAGE_SHIFT, DEV_SHARED); > > + > > + zimage += addr & ~PAGE_MASK; > > > > if (zimage[ZIMAGE_MAGIC_OFFSET/4] != ZIMAGE_MAGIC) > > + { > > + clear_fixmap(FIXMAP_MISC); > > return -EINVAL; > > + } > > > > start = zimage[ZIMAGE_START_OFFSET/4]; > > end = zimage[ZIMAGE_END_OFFSET/4]; > > > > clear_fixmap(FIXMAP_MISC); > > > > + if ( end > addr + size ) > > + return -EINVAL; > > Should this also check for start == 0 && end > size? Possibly ought to be checking for (end - start) > size which covers both? Looking at it now comparing addr + size with end seems a bit nonsensical since addr is where it is now and end is the end of where it would like to be loaded (or the size if start == 0, which is what has saved us so far). > > > int kernel_prepare(struct kernel_info *info) > > { > > int rc; > > > > - rc = kernel_try_zimage_prepare(info); > > + paddr_t start, size; > > + > > + if ( early_info.modules.nr_mods > 1 ) > > + panic("Cannot handle dom0 initrd yet\n"); > > + > > + if ( early_info.modules.nr_mods < 1 ) > > + { > > + printk("No boot modules found, trying flash\n"); > > + start = KERNEL_FLASH_ADDRESS; > > + size = KERNEL_FLASH_SIZE; > > + } > > + else > > + { > > + printk("Loading kernel from boot module 1\n"); > > + start = early_info.modules.module[1].start; > > + size = early_info.modules.module[1].size; > > Do we want (here or elsewhere) to check that start is page-aligned? I think kernel_try_zimage_prepare tries to do the right thing Although you've made me look and I suspect it is buggy if start is < sizeof(zimage header) from the end of a page. It should probably just use copy_from_paddr into a local buffer. Ian. _______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxx http://lists.xen.org/xen-devel
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |