- Download Diff

Summary:

10331 uts: gfx_private bitmap framebuffer implementation

Review Request #1417 — Created Feb. 1, 2019 and submitted Feb. 4, 2019, 12:37 p.m.

Information
Owner:	tsoome
Repository:	illumos-gate
Branch:
Bugs:	10331
Depends On:
Blocks:	1421
Commit:	8e9c37c...
Reviewers
Groups:	general
People:

Description

uts: gfx_private implement cls callback & visual_io v4
uts: gfx_private add alpha blending
uts: provide fb_info via fbgattr dev_specific array.
uts: identify FB types and set up gfx_private based on type.
uts: gfx_private should use setmode drm callback.

Testing Done

apply https://www.illumos.org/rb/r/1356/ and boot. Note that alpha blend needs update in tem to make text to appear on screen with 32-bit depth.

Issues

0
0
14
0
All Issues: 14

Description	From	Last Updated
nit: "simplified" (but if it's going to be removed later, no big deal)	jlevon	Feb. 1, 2019, 2:08 p.m.
Both args are used	jlevon	Feb. 1, 2019, 2:07 p.m.
(optional) replace with __unused?	jlevon	Feb. 1, 2019, 2:07 p.m.
Actually (sorry) it makes me wonder why these routines pass in the cmd anyway? It's always attach or detach at ...	jlevon	Feb. 1, 2019, 3:35 p.m.
this comment seems wrong?	jlevon	Feb. 1, 2019, 2:06 p.m.
nit missing space after cast	jlevon	Feb. 1, 2019, 2:02 p.m.
nit missing space after cast	jlevon	Feb. 1, 2019, 2:01 p.m.
ditto	jlevon	Feb. 1, 2019, 2:01 p.m.
ditto	jlevon	Feb. 1, 2019, 2:01 p.m.
odd formatting?	jlevon	Feb. 1, 2019, 1:19 p.m.
Nit: you should use this consistently (e.g. :1073) if you're going to have it	jlevon	Feb. 1, 2019, 1:16 p.m.
Why is this still here?	jlevon	Feb. 1, 2019, 1:14 p.m.
These get cleaned up later?	jlevon	Feb. 1, 2019, 1:14 p.m.
For outputting to error, "Cannot" might be preferable as the common form versus "Can not" -- though I also see ...	seeemef@mac.com	Feb. 3, 2019, 8:45 a.m.

Testing Done:

apply https://www.illumos.org/rb/r/1356/ and boot. Note that alpha blend needs update in tem to make text to appear on screen with 32-bit depth.

usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)

Show all issues

nit: "simplified" (but if it's going to be removed later, no big deal)

tsoome Feb. 1, 2019, 2:09 p.m.
```
Removed. it still is simple, however:)
```

usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
Both args are used
```
usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
(optional) replace with __unused?
```
usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
this comment seems wrong?
```
1. tsoome Feb. 1, 2019, 2:09 p.m.
  not entirely - I meant the fb content is in shadow_fb.
usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
nit missing space after cast
```
usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
nit missing space after cast
```
usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
ditto
```
usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revision 1)
Show all issues
```
ditto
```
usr/src/uts/i86pc/io/gfx_private/gfxp_fb.c (Diff revision 1)
Show all issues
```
odd formatting?
```

usr/src/uts/i86pc/io/gfx_private/gfxp_vgatext.c (Diff revision 1)

Show all issues

Nit: you should use this consistently (e.g. :1073) if you're going to have it

usr/src/uts/i86pc/io/gfx_private/gfxp_vgatext.c (Diff revision 1)
Show all issues
```
Why is this still here?
```

usr/src/uts/i86pc/io/gfx_private/gfxp_vgatext.c (Diff revision 1)

Show all issues

These get cleaned up later?

tsoome Feb. 1, 2019, 2:09 p.m.

Removed. I kept this stuff around for just a case, we probably should eventually remove it all unless someone wants to implement vga gfx mode.

Commit:

-	58de5a20bf803289896667d49b4f2659ad5d80e7
+	56550a3d67c7f757831866d33705631b1a8c4dea

Diff:

Revision 2 (+1511 -612)

Show changes

	usr/src/uts/common/sys/fbio.h
	usr/src/uts/common/sys/gfx_private.h
	usr/src/uts/i86pc/Makefile.files
	usr/src/uts/i86pc/io/consplat.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_fb.h
	usr/src/uts/i86pc/io/gfx_private/gfxp_fb.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_vgatext.c
	1 more

usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c (Diff revisions 1 - 2)

Show all issues

Actually (sorry) it makes me wonder why these routines pass in the cmd anyway? It's always attach or detach at this point?

tsoome Feb. 1, 2019, 3:59 p.m.
```
hm, yep, thats rigth.
```

Commit:

-	56550a3d67c7f757831866d33705631b1a8c4dea
+	3d0d2fbcbd03fee56147f77e1e8f671a566af6fc

Diff:

Revision 3 (+1509 -615)

Show changes

	usr/src/uts/common/sys/fbio.h
	usr/src/uts/common/sys/gfx_private.h
	usr/src/uts/i86pc/Makefile.files
	usr/src/uts/i86pc/io/consplat.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_fb.h
	usr/src/uts/i86pc/io/gfx_private/gfxp_fb.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_vgatext.c
	1 more

Ship it!

```
Ship It!
```

usr/src/uts/i86pc/io/gfx_private/gfxp_fb.c (Diff revision 3)

Show all issues

For outputting to error, "Cannot" might be preferable as the common form versus "Can not" -- though I also see below "can't" used with cmn_err().

tsoome Feb. 3, 2019, 8:48 a.m.

This code has been travelling from vgatext to gfxp_vgatext to there:)

Commit:

-	3d0d2fbcbd03fee56147f77e1e8f671a566af6fc
+	8e9c37c643359e6143376fea2fead7b0bf2fd5e9

Diff:

Revision 4 (+1509 -615)

Show changes

	usr/src/uts/common/sys/fbio.h
	usr/src/uts/common/sys/gfx_private.h
	usr/src/uts/i86pc/Makefile.files
	usr/src/uts/i86pc/io/consplat.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_bitmap.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_fb.h
	usr/src/uts/i86pc/io/gfx_private/gfxp_fb.c
	usr/src/uts/i86pc/io/gfx_private/gfxp_vgatext.c
	1 more

Ship it!

```
LGTM
```

You have a pending review.

Review Board 3.0.18

10331 uts: gfx_private bitmap framebuffer implementation

Testing Done:

Commit:

Diff:

Commit:

Diff:

Commit:

Diff:

Status: Closed (submitted)