dri-devel July 2018

dri-devel@lists.freedesktop.org

277 participants
1883 discussions

[PATCH v2 fbdev-for-next 1/2] fbcon: introduce for_each_registered_fb() helper
by Yisheng Xie 24 Jul '18

24 Jul '18

Following pattern is often used: for (i = 0; i < FB_MAX; i++) { if (registered_fb[i]) { ... } } Therefore, as Andy's suggestion, for_each_registered_fb() helper can be introduced to make the code easier to read and write by reducing indentation level. It also saves few lines of code in each occurrence. This patch convert all part here at the same time. Suggested-by: Andy Shevchenko <andriy.shevchenko(a)linux.intel.com> Signed-off-by: Yisheng Xie <ysxie(a)foxmail.com> --- v2: - rebase it to fbdev-for-next branch and add one more loop to replace - per Hans - Macro should be protected against nested conditionals. - per Andy drivers/video/fbdev/core/fbcon.c | 31 +++++++++++-------------------- drivers/video/fbdev/core/fbmem.c | 4 +--- include/linux/fb.h | 4 ++++ 3 files changed, 16 insertions(+), 23 deletions(-) diff --git a/drivers/video/fbdev/core/fbcon.c b/drivers/video/fbdev/core/fbcon.c index 5fb156b..e30d3a1 100644 --- a/drivers/video/fbdev/core/fbcon.c +++ b/drivers/video/fbdev/core/fbcon.c @@ -2234,8 +2234,8 @@ static int fbcon_switch(struct vc_data *vc) * * info->currcon = vc->vc_num; */ - for (i = 0; i < FB_MAX; i++) { - if (registered_fb[i] != NULL && registered_fb[i]->fbcon_par) { + for_each_registered_fb(i) { + if (registered_fb[i]->fbcon_par) { struct fbcon_ops *o = registered_fb[i]->fbcon_par; o->currcon = vc->vc_num; @@ -3124,11 +3124,9 @@ static int fbcon_fb_unregistered(struct fb_info *info) if (idx == info_idx) { info_idx = -1; - for (i = 0; i < FB_MAX; i++) { - if (registered_fb[i] != NULL) { - info_idx = i; - break; - } + for_each_registered_fb(i) { + info_idx = i; + break; } } @@ -3609,10 +3607,8 @@ static int fbcon_output_notifier(struct notifier_block *nb, deferred_takeover = false; logo_shown = FBCON_LOGO_DONTSHOW; - for (i = 0; i < FB_MAX; i++) { - if (registered_fb[i]) - fbcon_fb_registered(registered_fb[i]); - } + for_each_registered_fb(i) + fbcon_fb_registered(registered_fb[i]); return NOTIFY_OK; } @@ -3638,11 +3634,9 @@ static void fbcon_start(void) console_lock(); - for (i = 0; i < FB_MAX; i++) { - if (registered_fb[i] != NULL) { - info_idx = i; - break; - } + for_each_registered_fb(i) { + info_idx = i; + break; } do_fbcon_takeover(0); @@ -3669,15 +3663,12 @@ static void fbcon_exit(void) kfree((void *)softback_buf); softback_buf = 0UL; - for (i = 0; i < FB_MAX; i++) { + for_each_registered_fb(i) { int pending = 0; mapped = 0; info = registered_fb[i]; - if (info == NULL) - continue; - if (info->queue.func) pending = cancel_work_sync(&info->queue); DPRINTK("fbcon: %s pending work\n", (pending ? "canceled" : diff --git a/drivers/video/fbdev/core/fbmem.c b/drivers/video/fbdev/core/fbmem.c index 609438d..645c6ac 100644 --- a/drivers/video/fbdev/core/fbmem.c +++ b/drivers/video/fbdev/core/fbmem.c @@ -1593,10 +1593,8 @@ static int do_remove_conflicting_framebuffers(struct apertures_struct *a, int i, ret; /* check all firmware fbs and kick off if the base addr overlaps */ - for (i = 0 ; i < FB_MAX; i++) { + for_each_registered_fb(i) { struct apertures_struct *gen_aper; - if (!registered_fb[i]) - continue; if (!(registered_fb[i]->flags & FBINFO_MISC_FIRMWARE)) continue; diff --git a/include/linux/fb.h b/include/linux/fb.h index aa74a22..fd31e6f 100644 --- a/include/linux/fb.h +++ b/include/linux/fb.h @@ -650,6 +650,10 @@ extern int fb_get_color_depth(struct fb_var_screeninfo *var, extern int num_registered_fb; extern struct class *fb_class; +#define for_each_registered_fb(i) \ + for (i = 0; i < FB_MAX; i++) \ + if (!registered_fb[i]) {} else + extern int lock_fb_info(struct fb_info *info); static inline void unlock_fb_info(struct fb_info *info) -- 1.9.1

5 4

[PATCH] video: fbdev: fsl-diu-fb: Remove VLA usage
by Kees Cook 24 Jul '18

24 Jul '18

In the quest to remove all stack VLA usage from the kernel[1], this moves the buffer off the stack (since it could be as much as 1024 bytes), and uses a new area in the cursor data structure. Additionally adds missed documentation and removes redundant assignments. [1] https://lkml.kernel.org/r/CA+55aFzCG-zNmZwX4A2FQpadafLfEzK6CC=qPXydAacU1RqZ… Signed-off-by: Kees Cook <keescook(a)chromium.org> --- drivers/video/fbdev/fsl-diu-fb.c | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/drivers/video/fbdev/fsl-diu-fb.c b/drivers/video/fbdev/fsl-diu-fb.c index 1bfd13cbd4e3..bc9eb8afc313 100644 --- a/drivers/video/fbdev/fsl-diu-fb.c +++ b/drivers/video/fbdev/fsl-diu-fb.c @@ -360,6 +360,10 @@ struct mfb_info { * @ad[]: Area Descriptors for each real AOI * @gamma: gamma color table * @cursor: hardware cursor data + * @blank_cursor: blank cursor for hiding cursor + * @next_cursor: scratch space to build load cursor + * @edid_data: EDID information buffer + * @has_edid: whether or not the EDID buffer is valid * * This data structure must be allocated with 32-byte alignment, so that the * internal fields can be aligned properly. @@ -381,6 +385,8 @@ struct fsl_diu_data { __le16 cursor[MAX_CURS * MAX_CURS] __aligned(32); /* Blank cursor data -- used to hide the cursor */ __le16 blank_cursor[MAX_CURS * MAX_CURS] __aligned(32); + /* Scratch cursor data -- used to build new cursor */ + __le16 next_cursor[MAX_CURS * MAX_CURS] __aligned(32); uint8_t edid_data[EDID_LENGTH]; bool has_edid; } __aligned(32); @@ -1056,13 +1062,17 @@ static int fsl_diu_cursor(struct fb_info *info, struct fb_cursor *cursor) * FB_CUR_SETSHAPE - the cursor bitmask has changed */ if (cursor->set & (FB_CUR_SETSHAPE | FB_CUR_SETCMAP | FB_CUR_SETIMAGE)) { + /* + * Determine the size of the cursor image data. Normally, + * it's 8x16. + */ unsigned int image_size = - DIV_ROUND_UP(cursor->image.width, 8) * cursor->image.height; + DIV_ROUND_UP(cursor->image.width, 8) * + cursor->image.height; unsigned int image_words = DIV_ROUND_UP(image_size, sizeof(uint32_t)); unsigned int bg_idx = cursor->image.bg_color; unsigned int fg_idx = cursor->image.fg_color; - uint8_t buffer[image_size]; uint32_t *image, *source, *mask; uint16_t fg, bg; unsigned int i; @@ -1070,13 +1080,6 @@ static int fsl_diu_cursor(struct fb_info *info, struct fb_cursor *cursor) if (info->state != FBINFO_STATE_RUNNING) return 0; - /* - * Determine the size of the cursor image data. Normally, - * it's 8x16. - */ - image_size = DIV_ROUND_UP(cursor->image.width, 8) * - cursor->image.height; - bg = ((info->cmap.red[bg_idx] & 0xf8) << 7) | ((info->cmap.green[bg_idx] & 0xf8) << 2) | ((info->cmap.blue[bg_idx] & 0xf8) >> 3) | @@ -1088,7 +1091,7 @@ static int fsl_diu_cursor(struct fb_info *info, struct fb_cursor *cursor) 1 << 15; /* Use 32-bit operations on the data to improve performance */ - image = (uint32_t *)buffer; + image = (uint32_t *)data->next_cursor; source = (uint32_t *)cursor->image.data; mask = (uint32_t *)cursor->mask; -- 2.17.1 -- Kees Cook Pixel Security

3 2

Re: [PATCH 1/4] video: fbdev: pxafb: clear allocated memory for video modes
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Monday, July 09, 2018 07:12:50 AM Daniel Mack wrote: > Hi Bartlomiej, Hi, > Should I resend with Robert's Reviewed-bys again? I'd like to get this > merged for 4.19 if possible. No need for resend, I added Robert's tags while applying your patches. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

2 1

Re: [PATCH 4/4] video: fbdev: pxafb: Add support for lcd-supply regulator
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Sunday, June 24, 2018 05:38:17 PM Daniel Mack wrote: > Optionally obtain a lcd-supply regulator during probe and use it in > __pxafb_lcd_power() to switch the power supply of LCD panels. > > This helps boards booted from DT to control such voltages without > callbacks. > > Signed-off-by: Daniel Mack <daniel(a)zonque.org> Patch queued for 4.19, thanks. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Re: [PATCH 3/4] video: fbdev: pxafb: handle errors from pxafb_init_fbinfo() correctly
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Sunday, June 24, 2018 05:38:16 PM Daniel Mack wrote: > pxafb_init_fbinfo() can not only report errors caused by failed > allocations but also when the clock can't be found. > > To fix this, return an error pointer instead of NULL in case of errors, > and up-chain the result. > > Signed-off-by: Daniel Mack <daniel(a)zonque.org> Patch queued for 4.19, thanks. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Re: [PATCH 2/4] video: fbdev: pxafb: switch to devm_* API
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Sunday, June 24, 2018 05:38:15 PM Daniel Mack wrote: > This helps us clean up the probe() and remove() implementations. > > Signed-off-by: Daniel Mack <daniel(a)zonque.org> Patch queued for 4.19, thanks. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Re: [PATCH 1/4] video: fbdev: pxafb: clear allocated memory for video modes
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Sunday, June 24, 2018 05:38:14 PM Daniel Mack wrote: > When parsing the video modes from DT properties, make sure to zero out > memory before using it. This is important because not all fields in the mode > struct are explicitly initialized, even though they are used later on. > > Fixes: 420a488278e86 (video: fbdev: pxafb: initial devicetree conversion) > Reviewed-by: Robert Jarzmik <robert.jarzmik(a)free.fr> > Signed-off-by: Daniel Mack <daniel(a)zonque.org> Patch queued for 4.19, thanks. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Re: [PATCH v2 2/2] video: fbdev: pxa3xx_gcu: add devicetree bindings
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Wednesday, July 04, 2018 09:35:02 PM Daniel Mack wrote: > Add a device tree match table for this hardware graphics acceleration > driver so it can be used by pxa3xx boards booted with a devicetree. > > Signed-off-by: Daniel Mack <daniel(a)zonque.org> Patch queued for 4.19, thanks. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Re: [PATCH v2 1/2] dt-bindings: display: add devicetree bindings for pxa300-gcu
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Wednesday, July 04, 2018 09:35:01 PM Daniel Mack wrote: > This patch adds the binding documentation for the pxa300 gcu. > > Signed-off-by: Daniel Mack <daniel(a)zonque.org> Patch queued for 4.19, thanks. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Re: [PATCH] fbdev: Distinguish between interlaced and progressive modes
by Bartlomiej Zolnierkiewicz 24 Jul '18

24 Jul '18

On Wednesday, July 04, 2018 09:08:48 PM Fredrik Noring wrote: > Hi Bartlomiej, > > > > With this change fb_find_mode can match interlaced and progressive > > > modes equally well, and distinguish between for example 1920x1080i > > > and 1920x1080p. > > > > > > Signed-off-by: Fredrik Noring <noring(a)nocrew.org> > > > > Could you please give some examples of use cases that need this change > > (i.e. which fbdev drivers need it for which modes etc.)? > > I discovered the problem when developing a frame buffer driver for the > PlayStation 2 (not yet merged), using the following video modes for the > PlayStation 3 in drivers/video/fbdev/ps3fb.c: > > }, { > /* 1080if */ > "1080if", 50, 1920, 1080, 13468, 148, 484, 36, 4, 88, 5, > FB_SYNC_BROADCAST, FB_VMODE_INTERLACED > }, { > /* 1080pf */ > "1080pf", 50, 1920, 1080, 6734, 148, 484, 36, 4, 88, 5, > FB_SYNC_BROADCAST, FB_VMODE_NONINTERLACED > }, > > In ps3fb_probe, the mode_option module parameter is used with fb_find_mode > but it can only select the interlaced variant of 1920x1080 since the loop > matching the modes does not take the difference between interlaced and > progressive modes into account. > > In short, without the patch, progressive 1920x1080 cannot be chosen as a > mode_option parameter since fb_find_mode (falsely) thinks interlace is a > perfect match. Patch queued for 4.19 (with the above explanation used as patch description), thanks! Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel July 2018