dri-devel March 2021

dri-devel@lists.freedesktop.org

326 participants
867 discussions

[PATCH v4 0/3] Fixes to bridge/panel and ingenic-drm
by Paul Cercueil 29 Mar '21

29 Mar '21

Hi, This is the V4 of my patchset that attempts to fix use-after-free errors in bridge/panel.c and in the ingenic-drm driver. Changes from v3: - [1/3]: the code now checks (connector->dev) instead of (!!panel_bridge->connector.dev) - [2/3]: the macro is now called drmm_plain_encoder_alloc(), and moved to <drm/drm_encoder.h>. It also takes funcs/name parameters to be more similar to drmm_encoder_alloc(), although these parameters can very well be NULL. - [3/3] uses the new macro. V3 had a 4th patch, which was already applied as it received a reviewed-by tag and could be applied independently. Cheers, -Paul Paul Cercueil (3): drm: bridge/panel: Cleanup connector on bridge detach drm/encoder: Add macro drmm_plain_encoder_alloc() drm/ingenic: Register devm action to cleanup encoders drivers/gpu/drm/bridge/panel.c | 12 ++++++++++++ drivers/gpu/drm/ingenic/ingenic-drm-drv.c | 17 +++++++---------- include/drm/drm_encoder.h | 18 ++++++++++++++++++ 3 files changed, 37 insertions(+), 10 deletions(-) -- 2.30.2

2 5

[TRIVIAL] drm/amd/display: fix typo: liason -> liaison
by Diego Viola 29 Mar '21

29 Mar '21

Signed-off-by: Diego Viola <diego.viola(a)gmail.com> --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 573cf17262da..1b4b4f508662 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -121,7 +121,7 @@ MODULE_FIRMWARE(FIRMWARE_NAVI12_DMCU); * DOC: overview * * The AMDgpu display manager, **amdgpu_dm** (or even simpler, - * **dm**) sits between DRM and DC. It acts as a liason, converting DRM + * **dm**) sits between DRM and DC. It acts as a liaison, converting DRM * requests into DC requests, and DC responses into DRM responses. * * The root control structure is &struct amdgpu_display_manager. -- 2.31.1

2 1

[PATCH] drm/komeda: Fix off-by-1 when with readback conn due to rounding
by carsten.haitzler＠foss.arm.com 29 Mar '21

29 Mar '21

From: Carsten Haitzler <carsten.haitzler(a)arm.com> When setting up a readback connector that writes data back to memory rather than to an actual output device (HDMI etc.), rounding was set to round. As the DPU uses a higher internal number of bits when generating a color value, this round-down back to 8bit ended up with everything being off-by one. e.g. #fefefe became #ffffff. This sets rounding to "round-down" so things end up correct by turning on the LW_TRC round down flag. Signed-off-by: Carsten Haitzler <carsten.haitzler(a)arm.com> --- drivers/gpu/drm/arm/display/komeda/d71/d71_component.c | 7 ++++++- drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h | 1 + 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c index 8a02ade369db..e97acc5519d1 100644 --- a/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c @@ -468,7 +468,12 @@ static void d71_wb_layer_update(struct komeda_component *c, struct komeda_layer_state *st = to_layer_st(state); struct drm_connector_state *conn_st = state->wb_conn->state; struct komeda_fb *kfb = to_kfb(conn_st->writeback_job->fb); - u32 ctrl = L_EN | LW_OFM, mask = L_EN | LW_OFM | LW_TBU_EN; + /* LW_TRC sets rounding to truncate not round which is needed for + * the output of writeback to match the input in the most common + * use cases like RGB888 -> RGB888, so set this bit by default + */ + u32 ctrl = LW_TRC | L_EN | LW_OFM; + u32 mask = LW_TRC | L_EN | LW_OFM | LW_TBU_EN; u32 __iomem *reg = c->reg; d71_layer_update_fb(c, kfb, st->addr); diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h index e80172a0b320..a8036689d721 100644 --- a/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h @@ -321,6 +321,7 @@ #define LAYER_WR_FORMAT 0x0D8 /* Layer_WR control bits */ +#define LW_TRC BIT(1) #define LW_OFM BIT(4) #define LW_LALPHA(x) (((x) & 0xFF) << 8) #define LW_A_WCACHE(x) (((x) & 0xF) << 28) -- 2.30.0

4 3

[PATCH] i915: Fix uninitialized variable err
by Colin King 29 Mar '21

29 Mar '21

From: Colin Ian King <colin.king(a)canonical.com> In the case where !sg_dma_len(sgl) breaks out of the do-while loop on the first iteration, error variable err has not been assigned any value and will contain garbage. Fix this by ensuring err is initialized to zero. Addresses-Coverity: ("Uninitialized scalar variable") Fixes: 204302d90503 ("i915: fix remap_io_sg to verify the pgprot") Signed-off-by: Colin Ian King <colin.king(a)canonical.com> --- drivers/gpu/drm/i915/i915_mm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_mm.c b/drivers/gpu/drm/i915/i915_mm.c index 4c8cd08c672d..25576fa73ff0 100644 --- a/drivers/gpu/drm/i915/i915_mm.c +++ b/drivers/gpu/drm/i915/i915_mm.c @@ -47,7 +47,7 @@ int remap_io_sg(struct vm_area_struct *vma, struct scatterlist *sgl, resource_size_t iobase) { unsigned long pfn, len, remapped = 0; - int err; + int err = 0; /* We rely on prevalidation of the io-mapping to skip track_pfn(). */ GEM_BUG_ON((vma->vm_flags & EXPECTED_FLAGS) != EXPECTED_FLAGS); -- 2.30.2

1 0

[bug report] drm/amd/display: Reuse parsing code of debugfs write buffer
by Dan Carpenter 29 Mar '21

29 Mar '21

Hello Mikita Lipski, The patch 04111850cf56: "drm/amd/display: Reuse parsing code of debugfs write buffer" from Mar 26, 2020, leads to the following static checker warning: drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_debugfs.c:80 parse_write_buffer_into_params() error: 'copy_from_user()' 'wr_buf_ptr' too small (100 vs 1000000000) drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_debugfs.c 64 static int parse_write_buffer_into_params(char *wr_buf, uint32_t wr_buf_size, 65 long *param, const char __user *buf, 66 int max_param_num, 67 uint8_t *param_nums) 68 { 69 char *wr_buf_ptr = NULL; 70 uint32_t wr_buf_count = 0; 71 int r; 72 char *sub_str = NULL; 73 const char delimiter[3] = {' ', '\n', '\0'}; 74 uint8_t param_index = 0; 75 76 *param_nums = 0; 77 78 wr_buf_ptr = wr_buf; 79 80 r = copy_from_user(wr_buf_ptr, buf, wr_buf_size); ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The "wr_buf_ptr" is a fixed size buffer and the "wr_buf_size" comes from the user so this would be a privalege escalation bug if it weren't debugfs. It's better to move the allocation into this function. Remember to free before returning. wr_buf = memdup_user(buf, wr_buf_size + 1); if (IS_ERR(wr_buf)) return PTR_ERR(wr_buf); (Written in email client. Not compile tested.) 81 82 /* r is bytes not be copied */ 83 if (r >= wr_buf_size) { This test is wrong. It should have been zero instead of wr_buf_size but that doesn't matter now that we've switched to memdup_user(). 84 DRM_DEBUG_DRIVER("user data not be read\n"); 85 return -EINVAL; Don't print an error if copy_from_user() fails. Return -EFAULT instead of -EINVAL. 86 } 87 88 /* check number of parameters. isspace could not differ space and \n */ 89 while ((*wr_buf_ptr != 0xa) && (wr_buf_count < wr_buf_size)) { 90 /* skip space*/ 91 while (isspace(*wr_buf_ptr) && (wr_buf_count < wr_buf_size)) { 92 wr_buf_ptr++; 93 wr_buf_count++; 94 } 95 96 if (wr_buf_count == wr_buf_size) 97 break; 98 99 /* skip non-space*/ 100 while ((!isspace(*wr_buf_ptr)) && (wr_buf_count < wr_buf_size)) { 101 wr_buf_ptr++; 102 wr_buf_count++; 103 } 104 105 (*param_nums)++; 106 107 if (wr_buf_count == wr_buf_size) 108 break; 109 } 110 111 if (*param_nums > max_param_num) 112 *param_nums = max_param_num; 113 114 wr_buf_ptr = wr_buf; /* reset buf pointer */ 115 wr_buf_count = 0; /* number of char already checked */ 116 117 while (isspace(*wr_buf_ptr) && (wr_buf_count < wr_buf_size)) { 118 wr_buf_ptr++; 119 wr_buf_count++; 120 } 121 122 while (param_index < *param_nums) { 123 /* after strsep, wr_buf_ptr will be moved to after space */ 124 sub_str = strsep(&wr_buf_ptr, delimiter); This code is not checking wr_buf_count. I guess the thinking was that the first look would find the number of words and the next loop would use that pre-validated count. There is not a one to one relationship between your delimeters and ispace() (tabs) but I guess your delimeter is a subset of isspace() so that works... This code also assumes that "wr_buf_ptr" is NUL terminated but that's not necessarily the case so it could lead to a buffer overflow. For example, imagine the whole buffer is full of tabs. That's counted as one word. We're already off the end of the buffer before we call strsep(). Or if the whole buffer is spaces except for the last character. That's again one word and the strsep will read beyond the end and corrupt memory. 125 126 r = kstrtol(sub_str, 16, &(param[param_index])); 127 128 if (r) 129 DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r); 130 131 param_index++; 132 } 133 134 return 0; 135 } This whole function could probably be written like (not compiled): wr_buf = memdup_user(buf, wr_buf_size + 1); if (IS_ERR(wr_buf)) return PTR_ERR(wr_buf); wr_buf[wr_buf_size] = '\0'; p = wr_buf; while (p && *p) { if (*param_nums >= max_param_num) goto done; while (*p && isspace(*p)) p++; sub_str = strsep(&p, delimiter); ret = kstrtol(sub_str, 16, &(param[param_index])); if (ret) goto done; (*param_nums)++; } done: kfree(wr_buf); return 0; regards, dan carpenter

1 0

[PATCH v5 0/2] gpu: drm: add driver for ili9361 panel
by Daniel Mack 29 Mar '21

29 Mar '21

This is v3 of the series. Changelog: v2 -> v3: * Turn Documentation into yaml format v3 -> v4: * Fix reference error in yaml file v4 -> v5: * More yaml file documentation fixes Daniel Mack (2): dt-bindings: display: add bindings for newhaven,1.8-128160EF drm/tiny: add driver for newhaven,1.8-128160EF .../display/panel/ilitek,ili9163.yaml | 69 +++++++++++++++++++ drivers/gpu/drm/tiny/Kconfig | 13 + drivers/gpu/drm/tiny/Makefile | 1 + drivers/gpu/drm/tiny/ili9163.c | 224 ++++++++++++++++++ 4 files changed, 307 insertions(+) create mode 100644 Documentation/devicetree/bindings/display/ilitek,ili9163.yaml create mode 100644 drivers/gpu/drm/tiny/ili9163.c -- 2.29.2

2 3

Re: [PATCH 00/30] DMA: Mundane typo fixes
by Bhaskar Chowdhury 29 Mar '21

29 Mar '21

On 07:29 Mon 29 Mar 2021, Christoph Hellwig wrote: >I really don't think these typo patchbomb are that useful. I'm all >for fixing typos when working with a subsystem, but I'm not sure these >patchbombs help anything. > I am sure you are holding the wrong end of the wand and grossly failing to understand. Anyway, I hope I give a heads up ...find "your way" to fix those damn thing...it's glaring.... >On Mon, Mar 29, 2021 at 05:22:56AM +0530, Bhaskar Chowdhury wrote: >> This patch series fixes some trivial and rudimentary spellings in the COMMENT >> sections. >> >> Bhaskar Chowdhury (30): >> acpi-dma.c: Fix couple of typos >> altera-msgdma.c: Couple of typos fixed >> amba-pl08x.c: Fixed couple of typos >> bcm-sba-raid.c: Few typos fixed >> bcm2835-dma.c: Fix a typo >> idma64.c: Fix couple of typos >> iop-adma.c: Few typos fixed >> mv_xor.c: Fix a typo >> mv_xor.h: Fixed a typo >> mv_xor_v2.c: Fix a typo >> nbpfaxi.c: Fixed a typo >> of-dma.c: Fixed a typo >> s3c24xx-dma.c: Fix a typo >> Revert "s3c24xx-dma.c: Fix a typo" >> s3c24xx-dma.c: Few typos fixed >> st_fdma.h: Fix couple of typos >> ste_dma40_ll.h: Fix a typo >> tegra20-apb-dma.c: Fixed a typo >> xgene-dma.c: Few spello fixes >> at_hdmac.c: Quite a few spello fixes >> owl-dma.c: Fix a typo >> at_hdmac_regs.h: Couple of typo fixes >> dma-jz4780.c: Fix a typo >> Kconfig: Change Synopsys to Synopsis >> ste_dma40.c: Few spello fixes >> dw-axi-dmac-platform.c: Few typos fixed >> dpaa2-qdma.c: Fix a typo >> usb-dmac.c: Fix a typo >> edma.c: Fix a typo >> xilinx_dma.c: Fix a typo >> >> drivers/dma/Kconfig | 8 ++++---- >> drivers/dma/acpi-dma.c | 4 ++-- >> drivers/dma/altera-msgdma.c | 4 ++-- >> drivers/dma/amba-pl08x.c | 4 ++-- >> drivers/dma/at_hdmac.c | 14 +++++++------- >> drivers/dma/at_hdmac_regs.h | 4 ++-- >> drivers/dma/bcm-sba-raid.c | 8 ++++---- >> drivers/dma/bcm2835-dma.c | 2 +- >> drivers/dma/dma-jz4780.c | 2 +- >> drivers/dma/dw-axi-dmac/dw-axi-dmac-platform.c | 8 ++++---- >> drivers/dma/idma64.c | 4 ++-- >> drivers/dma/iop-adma.c | 6 +++--- >> drivers/dma/mv_xor.c | 2 +- >> drivers/dma/mv_xor.h | 2 +- >> drivers/dma/mv_xor_v2.c | 2 +- >> drivers/dma/nbpfaxi.c | 2 +- >> drivers/dma/of-dma.c | 2 +- >> drivers/dma/owl-dma.c | 2 +- >> drivers/dma/s3c24xx-dma.c | 6 +++--- >> drivers/dma/sh/shdmac.c | 2 +- >> drivers/dma/sh/usb-dmac.c | 2 +- >> drivers/dma/st_fdma.h | 4 ++-- >> drivers/dma/ste_dma40.c | 10 +++++----- >> drivers/dma/ste_dma40_ll.h | 2 +- >> drivers/dma/tegra20-apb-dma.c | 2 +- >> drivers/dma/ti/edma.c | 2 +- >> drivers/dma/xgene-dma.c | 6 +++--- >> drivers/dma/xilinx/xilinx_dma.c | 2 +- >> 28 files changed, 59 insertions(+), 59 deletions(-) >> >> -- >> 2.26.3 >---end quoted text---

2 2

[GIT PULL] exynos-drm-fixes
by Inki Dae 29 Mar '21

29 Mar '21

Hi Dave, Just one cleanup to drop unused header inclusion. Please kindly let me know if there is any problem. Thanks, Inki Dae The following changes since commit 09d78dde88ef95a27b54a6e450ee700ccabdf39d: Merge tag 'drm-msm-fixes-2021-02-25' of https://gitlab.freedesktop.org/drm/msm into drm-fixes (2021-03-26 13:04:17 +1000) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos tags/exynos-drm-fixes-for-v5.12-rc6 for you to fetch changes up to 6161a435c1910d07ee00cc25af010889010e1f08: drm/exynos/decon5433: Remove the unused include statements (2021-03-29 19:53:23 +0900) ---------------------------------------------------------------- Just one cleanup which drops of_gpio.h inclusion. - This header file isn't used anymore so drop it. ---------------------------------------------------------------- Tian Tao (1): drm/exynos/decon5433: Remove the unused include statements drivers/gpu/drm/exynos/exynos5433_drm_decon.c | 1 - 1 file changed, 1 deletion(-)

1 0

Re: Color mode exposed to user space?
by Alex Deucher 29 Mar '21

29 Mar '21

+ dri-devel I don't think it's currently exposed anywhere. Alex On Wed, Mar 24, 2021 at 5:11 AM Werner Sembach <wse(a)tuxedocomputers.com> wrote: > > Hello, > > is the information which color mode is currently in used for a display (RGB, YCbCr444, or YCbCr420) exposed to user space somewhere? > > If no: Where would be the best place to put code to expose it to sysfs? > > Thanks in advance, > > Werner Sembach > > > _______________________________________________ > amd-gfx mailing list > amd-gfx(a)lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/amd-gfx

3 2

[PATCH v1 1/2] gpu: host1x: Add back arm_iommu_detach_device()
by Dmitry Osipenko 28 Mar '21

28 Mar '21

The case of CONFIG_ARM_DMA_USE_IOMMU=y was found to be broken for host1x driver. Add back the workaround using arm_iommu_detach_device() as a temporary solution. Cc: stable(a)vger.kernel.org Fixes: af1cbfb9bf0f ("gpu: host1x: Support DMA mapping of buffers" Signed-off-by: Dmitry Osipenko <digetx(a)gmail.com> --- drivers/gpu/host1x/dev.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c index d0ebb70e2fdd..a2c09dca4eef 100644 --- a/drivers/gpu/host1x/dev.c +++ b/drivers/gpu/host1x/dev.c @@ -18,6 +18,10 @@ #include <trace/events/host1x.h> #undef CREATE_TRACE_POINTS +#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU) +#include <asm/dma-iommu.h> +#endif + #include "bus.h" #include "channel.h" #include "debug.h" @@ -232,6 +236,17 @@ static struct iommu_domain *host1x_iommu_attach(struct host1x *host) struct iommu_domain *domain = iommu_get_domain_for_dev(host->dev); int err; +#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU) + if (host->dev->archdata.mapping) { + struct dma_iommu_mapping *mapping = + to_dma_iommu_mapping(host->dev); + arm_iommu_detach_device(host->dev); + arm_iommu_release_mapping(mapping); + + domain = iommu_get_domain_for_dev(host->dev); + } +#endif + /* * We may not always want to enable IOMMU support (for example if the * host1x firewall is already enabled and we don't support addressing -- 2.30.2

1 1

← Newer
1
...
25
26
27
28
29
30
31
...
87
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel March 2021