| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <aS7qty8ABRjwQWaF@e142607>
Date: Tue, 2 Dec 2025 13:33:43 +0000
From: Liviu Dudau <liviu.dudau@....com>
To: Rahul Kumar <rk0006818@...il.com>
Cc: maarten.lankhorst@...ux.intel.com, mripard@...nel.org,
tzimmermann@...e.de, airlied@...il.com, simona@...ll.ch,
dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 3/3] drm/komeda: Convert logging in d71_dev.c to drm_*
with drm_device parameter
On Tue, Nov 25, 2025 at 12:15:28PM +0530, Rahul Kumar wrote:
> Hi Liviu,
Hi Rahul,
Sorry, you caught me with your new series just before going on holiday so
did not started the review. Since coming back I had some other internal
issues to take care of, I will have a look at your series next week when
I should have more time for it.
Best regards,
Liviu
>
> On Tue, Nov 18, 2025 at 4:32 PM Rahul Kumar <rk0006818@...il.com> wrote:
> >
> > Replace DRM_DEBUG() and DRM_ERROR() calls in
> > drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c with the
> > drm_dbg() and drm_err() helpers in functions where a drm_device
> > parameter is available.
> >
> > The drm_*() logging macros require a struct drm_device * parameter,
> > which allows the DRM core to prefix log messages with the device
> > instance. This improves debugging clarity when multiple Komeda or
> > other DRM devices are present.
> >
> > Logging in early hardware probing functions such as d71_identify()
> > is intentionally left unchanged because they do not have access to
> > a drm_device pointer at that stage of initialization.
> >
> > This conversion follows the DRM TODO entry:
> > "Convert logging to drm_* functions with drm_device parameter".
> >
> > Signed-off-by: Rahul Kumar <rk0006818@...il.com>
> > ---
> > Changes since v1:
> > - Fixed incorrect use of dev_get_drvdata(): it returns struct komeda_drv *,
> > not struct komeda_kms_dev *, as pointed out by Liviu Dudau.
> > - Updated DRM device pointer retrieval to use drv->kms.
> > - Combined both changes into a 0/3 series.
> > Link to v1:
> > https://lore.kernel.org/all/aRdT1qscQqO7-U6h@e110455-lin.cambridge.arm.com/
> > ---
> > .../gpu/drm/arm/display/komeda/d71/d71_dev.c | 24 +++++++++++++------
> > 1 file changed, 17 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
> > index 80973975bfdb..39c51bbe2bb9 100644
> > --- a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
> > +++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
> > @@ -9,6 +9,7 @@
> > #include <drm/drm_print.h>
> > #include "d71_dev.h"
> > #include "malidp_io.h"
> > +#include "komeda_drv.h"
> >
> > static u64 get_lpu_event(struct d71_pipeline *d71_pipeline)
> > {
> > @@ -348,6 +349,9 @@ static void d71_cleanup(struct komeda_dev *mdev)
> >
> > static int d71_enum_resources(struct komeda_dev *mdev)
> > {
> > + struct komeda_drv *drv = dev_get_drvdata(mdev->dev);
> > + struct komeda_kms_dev *kms = drv->kms;
> > + struct drm_device *drm = &kms->base;
> > struct d71_dev *d71;
> > struct komeda_pipeline *pipe;
> > struct block_header blk;
> > @@ -366,7 +370,7 @@ static int d71_enum_resources(struct komeda_dev *mdev)
> >
> > err = d71_reset(d71);
> > if (err) {
> > - DRM_ERROR("Fail to reset d71 device.\n");
> > + drm_err(drm, "Fail to reset d71 device.\n");
> > goto err_cleanup;
> > }
> >
> > @@ -376,8 +380,8 @@ static int d71_enum_resources(struct komeda_dev *mdev)
> > d71->num_pipelines = (value >> 8) & 0x7;
> >
> > if (d71->num_pipelines > D71_MAX_PIPELINE) {
> > - DRM_ERROR("d71 supports %d pipelines, but got: %d.\n",
> > - D71_MAX_PIPELINE, d71->num_pipelines);
> > + drm_err(drm, "d71 supports %d pipelines, but got: %d.\n",
> > + D71_MAX_PIPELINE, d71->num_pipelines);
> > err = -EINVAL;
> > goto err_cleanup;
> > }
> > @@ -455,8 +459,8 @@ static int d71_enum_resources(struct komeda_dev *mdev)
> > offset += D71_BLOCK_SIZE;
> > }
> >
> > - DRM_DEBUG("total %d (out of %d) blocks are found.\n",
> > - i, d71->num_blocks);
> > + drm_dbg(drm, "total %d (out of %d) blocks are found.\n",
> > + i, d71->num_blocks);
> >
> > return 0;
> >
> > @@ -555,6 +559,9 @@ static void d71_init_fmt_tbl(struct komeda_dev *mdev)
> >
> > static int d71_connect_iommu(struct komeda_dev *mdev)
> > {
> > + struct komeda_drv *drv = dev_get_drvdata(mdev->dev);
> > + struct komeda_kms_dev *kms = drv->kms;
> > + struct drm_device *drm = &kms->base;
> > struct d71_dev *d71 = mdev->chip_data;
> > u32 __iomem *reg = d71->gcu_addr;
> > u32 check_bits = (d71->num_pipelines == 2) ?
> > @@ -569,7 +576,7 @@ static int d71_connect_iommu(struct komeda_dev *mdev)
> > ret = dp_wait_cond(has_bits(check_bits, malidp_read32(reg, BLK_STATUS)),
> > 100, 1000, 1000);
> > if (ret < 0) {
> > - DRM_ERROR("timed out connecting to TCU!\n");
> > + drm_err(drm, "timed out connecting to TCU!\n");
> > malidp_write32_mask(reg, BLK_CONTROL, 0x7, INACTIVE_MODE);
> > return ret;
> > }
> > @@ -582,6 +589,9 @@ static int d71_connect_iommu(struct komeda_dev *mdev)
> >
> > static int d71_disconnect_iommu(struct komeda_dev *mdev)
> > {
> > + struct komeda_drv *drv = dev_get_drvdata(mdev->dev);
> > + struct komeda_kms_dev *kms = drv->kms;
> > + struct drm_device *drm = &kms->base;
> > struct d71_dev *d71 = mdev->chip_data;
> > u32 __iomem *reg = d71->gcu_addr;
> > u32 check_bits = (d71->num_pipelines == 2) ?
> > @@ -593,7 +603,7 @@ static int d71_disconnect_iommu(struct komeda_dev *mdev)
> > ret = dp_wait_cond(((malidp_read32(reg, BLK_STATUS) & check_bits) == 0),
> > 100, 1000, 1000);
> > if (ret < 0) {
> > - DRM_ERROR("timed out disconnecting from TCU!\n");
> > + drm_err(drm, "timed out disconnecting from TCU!\n");
> > malidp_write32_mask(reg, BLK_CONTROL, 0x7, INACTIVE_MODE);
> > }
> >
> > --
> > 2.43.0
>
>
> Just following up on my v2 Komeda logging patch series.
> Please let me know if any changes or suggestions are needed.
>
> Thanks,
> Rahul
Powered by blists - more mailing lists