Bug#588509: linux-image-2.6.32-5-amd64: hangs under Xen after Write protecting the kernel read-only data: 4220k

[Timo Juhani Lindfors <timo.lindfors@iki.fi>]

I changed

static atomic64_t last_value = ATOMIC64_INIT(0);

to

volatile static atomic64_t last_value = ATOMIC64_INIT(0);

but then Linux got stuck in

(gdb) bt 8
#0  panic (fmt=0x0) at /local/lindi/debian-linux-2.6/linux-2.6-2.6.32/debian/build/source_amd64_none/kernel/panic.c:137
#1  0xffffffff812e550f in find_early_table_space (use_gbpages=<value optimized out>, use_pse=<value optimized out>, end=<value optimized out>)
    at /local/lindi/debian-linux-2.6/linux-2.6-2.6.32/debian/build/source_amd64_none/arch/x86/mm/init.c:80
#2  init_memory_mapping (use_gbpages=<value optimized out>, use_pse=<value optimized out>, end=<value optimized out>)
    at /local/lindi/debian-linux-2.6/linux-2.6-2.6.32/debian/build/source_amd64_none/arch/x86/mm/init.c:271
#3  0xffffffff814f0e25 in ?? ()
#4  0xffffffff814ed9d0 in start_kernel () at /local/lindi/debian-linux-2.6/linux-2.6-2.6.32/debian/build/source_amd64_none/init/main.c:546
#5  0xffffffff8307f000 in ?? ()
#6  0xffffffff83080000 in ?? ()
#7  0xffffffff83081000 in ?? ()
(More stack frames follow...)

without printing anything. The check in question is probably

e820_table_start = find_e820_area(start, max_pfn_mapped<<PAGE_SHIFT,
                                tables, PAGE_SIZE);
if (e820_table_start == -1UL)
        panic("Cannot find space for the kernel page tables");


Any idea what is going on?


-Timo