Hi Am 12.01.24 um 21:38 schrieb Ian Forbes:
SVGA requires surfaces to fit within graphics memory (max_mob_pages) which means that modes with a final buffer size that would exceed graphics memory must be pruned otherwise creation will fail. Additionally, device commands which use multiple graphics resources must have all their resources fit within graphics memory for the duration of the command. Thus we need a small carve out of 1/4 of graphics memory to ensure commands likes surface copies to the primary framebuffer for cursor composition or damage clips can fit within graphics memory. This fixes an issue where VMs with low graphics memory (< 64MiB) configured with high resolution mode boot to a black screen because surface creation fails.
That is a long-standing problem, which we have observed with other drivers as well. On low-memory devices, TTM doesn't play well. The real fix would be to export all modes that possibly fit and sort out the invalid configurations in atomic_check. It's just a lot more work.
Did you consider simply ignoring vmwgfx devices with less than 64 MiB of VRAM?
Signed-off-by: Ian Forbes <ian.forbes@xxxxxxxxxxxx> --- drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c index 28ff30e32fab..39d6d17fc488 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c @@ -2854,12 +2854,12 @@ enum drm_mode_status vmw_connector_mode_valid(struct drm_connector *connector, struct vmw_private *dev_priv = vmw_priv(dev); u32 max_width = dev_priv->texture_max_width; u32 max_height = dev_priv->texture_max_height; - u32 assumed_cpp = 4; + u32 assumed_cpp = dev_priv->assume_16bpp ? 2 : 4; + u32 pitch = mode->hdisplay * assumed_cpp; + u64 total = mode->vdisplay * pitch; + bool using_stdu = dev_priv->active_display_unit == vmw_du_screen_target;- if (dev_priv->assume_16bpp)- assumed_cpp = 2; - - if (dev_priv->active_display_unit == vmw_du_screen_target) { + if (using_stdu) { max_width = min(dev_priv->stdu_max_width, max_width); max_height = min(dev_priv->stdu_max_height, max_height); } @@ -2870,9 +2870,13 @@ enum drm_mode_status vmw_connector_mode_valid(struct drm_connector *connector, if (max_height < mode->vdisplay) return MODE_BAD_VVALUE;- if (!vmw_kms_validate_mode_vram(dev_priv,- mode->hdisplay * assumed_cpp, - mode->vdisplay)) + if (using_stdu && + (total > (dev_priv->max_mob_pages * PAGE_SIZE * 3 / 4) ||
IDK the details of vmwgfx's memory management, but instead of this '3/4 test', wouldn't it be better to partition the VRAM via TTM and test against the partition sizes?
Best regards Thomas
+ total > dev_priv->max_mob_size)) {
+ return MODE_MEM;
+ }
+
+ if (!vmw_kms_validate_mode_vram(dev_priv, pitch, mode->vdisplay))
return MODE_MEM;
return MODE_OK;
-- 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
