| Message ID | 1456959362-2036-5-git-send-email-ddaney.cavm@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Mar 2, 2016 at 4:56 PM, David Daney <ddaney.cavm@gmail.com> wrote: > From: David Daney <david.daney@cavium.com> > > In order to extract NUMA information from the device tree, we need to > have the tree in its unflattened form. > > Move the call to unflatten_device_tree() into paging_init(). This > puts it before the call to bootmem_init(), which is where the NUMA > information is extracted. Can't you just move up unflatten_device_tree in setup_arch rather than hiding in paging_init? > Signed-off-by: David Daney <david.daney@cavium.com> > --- > arch/arm64/kernel/setup.c | 7 +++---- > arch/arm64/mm/mmu.c | 5 +++++ > 2 files changed, 8 insertions(+), 4 deletions(-)
On 03/03/2016 05:47 AM, Rob Herring wrote: > On Wed, Mar 2, 2016 at 4:56 PM, David Daney <ddaney.cavm@gmail.com> wrote: >> From: David Daney <david.daney@cavium.com> >> >> In order to extract NUMA information from the device tree, we need to >> have the tree in its unflattened form. >> >> Move the call to unflatten_device_tree() into paging_init(). This >> puts it before the call to bootmem_init(), which is where the NUMA >> information is extracted. > > Can't you just move up unflatten_device_tree in setup_arch rather than > hiding in paging_init? No. It must come *after* map_mem() and *before* bootmem_init(), both of which are done within paging_init(). One option would be to split the things in paging_init() into two functions, and then do: . . . paging_init_first_part(); if (acpi_disabled) unflatten_device_tree(); paging_init_second_part(); . . . > >> Signed-off-by: David Daney <david.daney@cavium.com> >> --- >> arch/arm64/kernel/setup.c | 7 +++---- >> arch/arm64/mm/mmu.c | 5 +++++ >> 2 files changed, 8 insertions(+), 4 deletions(-)
On 3 March 2016 at 17:33, David Daney <ddaney@caviumnetworks.com> wrote: > On 03/03/2016 05:47 AM, Rob Herring wrote: >> >> On Wed, Mar 2, 2016 at 4:56 PM, David Daney <ddaney.cavm@gmail.com> wrote: >>> >>> From: David Daney <david.daney@cavium.com> >>> >>> In order to extract NUMA information from the device tree, we need to >>> have the tree in its unflattened form. >>> >>> Move the call to unflatten_device_tree() into paging_init(). This >>> puts it before the call to bootmem_init(), which is where the NUMA >>> information is extracted. >> >> >> Can't you just move up unflatten_device_tree in setup_arch rather than >> hiding in paging_init? > > > No. > > It must come *after* map_mem() and *before* bootmem_init(), both of which > are done within paging_init(). > > One option would be to split the things in paging_init() into two functions, > and then do: > > . > . > . > paging_init_first_part(); > if (acpi_disabled) > unflatten_device_tree(); > paging_init_second_part(); > . I think it makes more sense to move bootmem_init() to setup_arch() then
diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 8119479..feae073 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -328,12 +328,11 @@ void __init setup_arch(char **cmdline_p) early_ioremap_reset(); - if (acpi_disabled) { - unflatten_device_tree(); + if (acpi_disabled) psci_dt_init(); - } else { + else psci_acpi_init(); - } + xen_early_init(); cpu_read_bootcpu_ops(); diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c index 58faeaa..ee6e6b0 100644 --- a/arch/arm64/mm/mmu.c +++ b/arch/arm64/mm/mmu.c @@ -29,6 +29,8 @@ #include <linux/io.h> #include <linux/slab.h> #include <linux/stop_machine.h> +#include <linux/acpi.h> +#include <linux/of_fdt.h> #include <asm/cputype.h> #include <asm/fixmap.h> @@ -459,6 +461,9 @@ void __init paging_init(void) map_mem(); fixup_executable(); + if (acpi_disabled) + unflatten_device_tree(); + /* allocate the zero page. */ zero_page = early_alloc(PAGE_SIZE);