[PATCH] nvn: Place flush_page below 4GB.

[Andi Kleen <ak@linux.intel.com>]

A system with an older Nvidia card keeps spewing this warning at boot
about NVN trying to DMA map the flush page that exceeds the 4GB mask of the
device.

I don't see any negative effects, so maybe the mask is incorrect,
but anyways, can just easily put the flush page into DMA32.

[    2.389824] nouveau 0000:65:00.0: bios: version 70.18.5f.00.00
[    2.389970] ------------[ cut here ]------------
[    2.389971] nouveau 0000:65:00.0: DMA addr 0x0000000118d1a000+4096 overflow (mask ffffffff, bus limit 0).
[    2.389977] WARNING: CPU: 0 PID: 8 at kernel/dma/direct.h:103 dma_map_page_attrs+0x280/0x2a0
[    2.389981] Modules linked in: nouveau(+) drm_ttm_helper ttm video gpu_sched crct10dif_pclmul crc32_pclmul i2c_algo_bit crc32c_intel drm_gpuvm polyval_clmulni drm_exec polyval_generic drm_display_helper mxm_wmi ghash_clmulni_intel sha512_ssse3 sha256_ssse3 cec sha1_ssse3 wdat_wdt wmi scsi_dh_rdac scsi_dh_emc scsi_dh_alua fuse
[    2.389996] CPU: 0 UID: 0 PID: 8 Comm: kworker/0:0 Not tainted 6.11.10-300.fc41.x86_64 #1
[    2.389998] Hardware name: Gigabyte Technology Co., Ltd. X299 AORUS Gaming 3 Pro/X299 AORUS Gaming 3 Pro-CF, BIOS F3 12/28/2017
[    2.389999] Workqueue: events work_for_cpu_fn
[    2.390003] RIP: 0010:dma_map_page_attrs+0x280/0x2a0
[    2.390005] Code: 8b 5d 00 48 89 ef e8 cf 7d 9d 00 4d 89 e1 4d 89 e8 48 89 da 48 8d 4c 24 08 41 56 48 89 c6 48 c7 c7 a0 30 be bb e8 00 d1 f2 ff <0f> 0b 58 eb 81 49 8b 43 08 83 78 68 05 0f 85 36 ff ff ff e9 6e ff
[    2.390006] RSP: 0018:ffffaca2c00e7c48 EFLAGS: 00010286
[    2.390008] RAX: 0000000000000000 RBX: ffff9bd044d2af20 RCX: 0000000000000027
[    2.390009] RDX: ffff9bdf9f021908 RSI: 0000000000000001 RDI: ffff9bdf9f021900
[    2.390010] RBP: ffff9bd0452e80c8 R08: 0000000000000000 R09: 202c666666666666
[    2.390011] R10: 6c20737562202c66 R11: 2e29302074696d69 R12: 00000000ffffffff
[    2.390012] R13: 0000000000001000 R14: 0000000000000000 R15: ffff9bd056d32c00
[    2.390013] FS:  0000000000000000(0000) GS:ffff9bdf9f000000(0000) knlGS:0000000000000000
[    2.390014] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    2.390015] CR2: 000055eba91b7000 CR3: 0000000af942a006 CR4: 00000000003706f0
[    2.390016] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    2.390017] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    2.390018] Call Trace:
[    2.390019]  <TASK>
[    2.390020]  ? dma_map_page_attrs+0x280/0x2a0
[    2.390022]  ? __warn.cold+0x8e/0xe8
[    2.390024]  ? dma_map_page_attrs+0x280/0x2a0
[    2.390026]  ? report_bug+0xff/0x140
[    2.390029]  ? console_unlock+0x84/0x130
[    2.390031]  ? handle_bug+0x58/0x90
[    2.390034]  ? exc_invalid_op+0x17/0x70
[    2.390035]  ? asm_exc_invalid_op+0x1a/0x20
[    2.390037]  ? dma_map_page_attrs+0x280/0x2a0
[    2.390038]  ? dma_map_page_attrs+0x280/0x2a0
[    2.390040]  nvkm_fb_ctor+0xba/0xe0 [nouveau]
[    2.390215]  nv50_fb_new_+0x91/0xc0 [nouveau]
[    2.390363]  nvkm_device_ctor+0x2457/0x4640 [nouveau]
[    2.390535]  nvkm_device_pci_new+0x12c/0x2f0 [nouveau]
[    2.390706]  nouveau_drm_probe+0xd2/0x220 [nouveau]
[    2.390870]  local_pci_probe+0x42/0x90
[    2.390874]  work_for_cpu_fn+0x17/0x30
[    2.390876]  process_one_work+0x176/0x330
[    2.390879]  worker_thread+0x252/0x390
[    2.390882]  ? __pfx_worker_thread+0x10/0x10
[    2.390884]  kthread+0xcf/0x100
[    2.390886]  ? __pfx_kthread+0x10/0x10
[    2.390888]  ret_from_fork+0x31/0x50
[    2.390891]  ? __pfx_kthread+0x10/0x10
[    2.390892]  ret_from_fork_asm+0x1a/0x30
[    2.390896]  </TASK>
[    2.390897] ---[ end trace 0000000000000000 ]---
[    2.391160] nouveau 0000:65:00.0: fb: 512 MiB DDR3
[    2.391165] nouveau 0000:65:00.0: fb: memory controller reports 512 MiB VRAM
[    3.099923] nouveau 0000:65:00.0: DRM: VRAM: 512 MiB
[    3.099925] nouveau 0000:65:00.0: DRM: GART: 1048576 MiB
[    3.099928] nouveau 0000:65:00.0: DRM: TMDS table version 2.0
[    3.101857] nouveau 0000:65:00.0: DRM: failed to create ce channel, -22
[    3.101980] nouveau 0000:65:00.0: DRM: MM: using COPY for buffer copies
[    3.102855] [drm] Initialized nouveau 1.4.0 for 0000:65:00.0 on minor 0
[    3.122956] nouveau 0000:65:00.0: [drm] Cannot find any crtc or sizes
[    3.165323] nouveau 0000:65:00.0: [drm] Cannot find any crtc or sizes
[    3.185419] nouveau 0000:65:00.0: [drm] Cannot find any crtc or sizes
[    3.206448] nouveau 0000:65:00.0: [drm] Cannot find any crtc or sizes
[    3.226508] nouveau 0000:65:00.0: [drm] Cannot find any crtc or sizes

Signed-off-by: Andi Kleen <ak@linux.intel.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c
index 8a286a9349ac..7b21ef375114 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c
@@ -279,7 +279,7 @@ nvkm_fb_ctor(const struct nvkm_fb_func *func, struct nvkm_device *device,
 	mutex_init(&fb->tags.mutex);
 
 	if (func->sysmem.flush_page_init) {
-		fb->sysmem.flush_page = alloc_page(GFP_KERNEL | __GFP_ZERO);
+		fb->sysmem.flush_page = alloc_page(GFP_KERNEL | __GFP_ZERO | GFP_DMA32);
 		if (!fb->sysmem.flush_page)
 			return -ENOMEM;
 
-- 
2.47.1