| Message ID | aabc5bad008d49f07d542815c6c8d28ec90bb09e.1628672091.git.geert+renesas@glider.be (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | memblock: Check memory add/cap ordering | expand |
On 11.08.21 10:55, Geert Uytterhoeven wrote: > For memblock_cap_memory_range() to work properly, it should be called > after memory is detected and added to memblock with memblock_add() or > memblock_add_node(). If memblock_cap_memory_range() would be called > before memory is registered, we may silently corrupt memory later > because the crash kernel will see all memory as available. > > Print a warning and bail out if ordering is not satisfied. > > Suggested-by: Mike Rapoport <rppt@kernel.org> > Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> > --- > mm/memblock.c | 5 +++++ > 1 file changed, 5 insertions(+) > > diff --git a/mm/memblock.c b/mm/memblock.c > index 57a9849a5d820c34..e2ca8ddc8ebebf4e 100644 > --- a/mm/memblock.c > +++ b/mm/memblock.c > @@ -1685,6 +1685,11 @@ void __init memblock_cap_memory_range(phys_addr_t base, phys_addr_t size) > if (!size) > return; > > + if (memblock.memory.cnt <= 1) { > + pr_warn("%s: No memory registered yet\n", __func__); > + return; > + } > + > ret = memblock_isolate_range(&memblock.memory, base, size, > &start_rgn, &end_rgn); > if (ret) > Reviewed-by: David Hildenbrand <david@redhat.com>
On Wed, Aug 11, 2021 at 10:55:18AM +0200, Geert Uytterhoeven wrote: > For memblock_cap_memory_range() to work properly, it should be called > after memory is detected and added to memblock with memblock_add() or > memblock_add_node(). If memblock_cap_memory_range() would be called > before memory is registered, we may silently corrupt memory later > because the crash kernel will see all memory as available. > > Print a warning and bail out if ordering is not satisfied. > > Suggested-by: Mike Rapoport <rppt@kernel.org> > Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Applied for 5.15, Thanks!
diff --git a/mm/memblock.c b/mm/memblock.c index 57a9849a5d820c34..e2ca8ddc8ebebf4e 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -1685,6 +1685,11 @@ void __init memblock_cap_memory_range(phys_addr_t base, phys_addr_t size) if (!size) return; + if (memblock.memory.cnt <= 1) { + pr_warn("%s: No memory registered yet\n", __func__); + return; + } + ret = memblock_isolate_range(&memblock.memory, base, size, &start_rgn, &end_rgn); if (ret)
For memblock_cap_memory_range() to work properly, it should be called after memory is detected and added to memblock with memblock_add() or memblock_add_node(). If memblock_cap_memory_range() would be called before memory is registered, we may silently corrupt memory later because the crash kernel will see all memory as available. Print a warning and bail out if ordering is not satisfied. Suggested-by: Mike Rapoport <rppt@kernel.org> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> --- mm/memblock.c | 5 +++++ 1 file changed, 5 insertions(+)