2003-06-24 17:05:29

by Adrian Bunk

Subject: [2.5 patch] small cleanups for amd-k8-agp.c

Hi Dave,

the patch below does the following trivial cleanups to
drivers/char/agp/amd-k8-agp.c:
- postfix three constants with ULL
- remove a variable that isn't needed

Please apply
Adrian


--- linux-2.5.73-not-full/drivers/char/agp/amd-k8-agp.c.old 2003-06-23 22:26:27.000000000 +0200
+++ linux-2.5.73-not-full/drivers/char/agp/amd-k8-agp.c 2003-06-23 22:27:21.000000000 +0200
@@ -47,7 +47,6 @@
static int x86_64_insert_memory(struct agp_memory *mem, off_t pg_start, int type)
{
int i, j, num_entries;
- long tmp;
u32 pte;
u64 addr;

@@ -78,10 +77,9 @@
for (i = 0, j = pg_start; i < mem->page_count; i++, j++) {
addr = agp_bridge->driver->mask_memory(mem->memory[i], mem->type);

- tmp = addr;
- BUG_ON(tmp & 0xffffff0000000ffc);
- pte = (tmp & 0x000000ff00000000) >> 28;
- pte |=(tmp & 0x00000000fffff000);
+ BUG_ON(addr & 0xffffff0000000ffcULL);
+ pte = (addr & 0x000000ff00000000ULL) >> 28;
+ pte |=(addr & 0x00000000fffff000ULL);
pte |= 1<<1|1<<0;

agp_bridge->gatt_table[j] = pte;

2003-06-26 00:44:27

by Dave Jones

Subject: Re: [2.5 patch] small cleanups for amd-k8-agp.c

On Tue, Jun 24, 2003 at 07:19:26PM +0200, Adrian Bunk wrote:

> the patch below does the following trivial cleanups to
> drivers/char/agp/amd-k8-agp.c:
> - postfix three constants with ULL
> - remove a variable that isn't needed
>
> Please apply

Rejected. That whole block of code is bogus when run in 32 bit mode.
The ULL warnings that are currently emitted by gcc serve as reminders for
me to get off my arse and fix it some day.

Dave