Hi Am 02.11.23 um 03:29 schrieb Peng Hao:
Use Macro DRM_FORMAT_MAX_PLANES instead of 4, to improve modifiability. Signed-off-by: Peng Hao <penghao@xxxxxxxxxxx> --- drivers/gpu/drm/drm_framebuffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_framebuffer.c b/drivers/gpu/drm/drm_framebuffer.c index 2dd97473ca10..bf283dae9090 100644 --- a/drivers/gpu/drm/drm_framebuffer.c +++ b/drivers/gpu/drm/drm_framebuffer.c @@ -254,7 +254,7 @@ static int framebuffer_check(struct drm_device *dev, } }- for (i = info->num_planes; i < 4; i++) {+ for (i = info->num_planes; i < DRM_FORMAT_MAX_PLANES; i++) {
This change makes the code more fragile. '4' is a fixed constant in the UAPI struct, while DRM_FORMAT_MAX_PLANES is an internal constant. I agree that both should reasonably have the same value. But (potentially) changing the value of DRM_FORMAT_MAX_PLANES will break these loops with a possible OOB access.
To make make this code more robust, it might be better to rewrite the tests like this
for (i = num_planes; i < ARRAY_SIZE(r->modifier); +i) {
// the test for modifier[i]
}
if (r->flags & DRM_MODE_FB_MODIFIERS) {
for (i < ARRAY_SIZE(handles)) {
// test for handles[i]
}
for (i < ARRAY_SIZE(pitches)) {
// test for pitches[i]
}
for (i < ARRAY_SIZE(offsets)) {
// test for offsets[i]
}
}
Best regards
Thomas
if (r->modifier[i]) {
drm_dbg_kms(dev, "non-zero modifier for unused plane %d\n", i);
return -EINVAL;
-- Thomas Zimmermann Graphics Driver Developer SUSE Software Solutions Germany GmbH Frankenstrasse 146, 90461 Nuernberg, Germany GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman HRB 36809 (AG Nuernberg)
Attachment:
OpenPGP_signature.asc
Description: OpenPGP digital signature
