Skip to contentSkip to conversation

platform/chrome: chromeos_acpi: print hex string for ACPI_TYPE_BUFFER

PATCH Superseded Tzung-Bi Shih 3 participants 8 messages Aug 2, 2023 – Aug 10, 2023

Subsystems (2)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Tzung-Bi Shih

author maintainer
#0 lore raw

PATCH v2 1/1 platform/chrome: chromeos_acpi: print hex string for ACPI_TYPE_BUFFER

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

---
Changes from v1[1]:
- Use hex_dump_to_buffer().
- Rewrite the loop for catching edge cases for truncating.

[1]: https://patchwork.kernel.org/project/chrome-platform/patch/[email protected]/

 drivers/platform/chrome/chromeos_acpi.c | 31 ++++++++++++++++++++++++-
 1 file changed, 30 insertions(+), 1 deletion(-)
drivers/platform/chrome/chromeos_acpi.c +30 -1
index 50d8a4d4352d..761506d307ad 100644
@@ -90,7 +90,36 @@ static int chromeos_acpi_handle_package(struct device *dev, union acpi_object *o
 	case ACPI_TYPE_STRING:
 		return sysfs_emit(buf, "%s\n", element->string.pointer);
 	case ACPI_TYPE_BUFFER:
-		return sysfs_emit(buf, "%s\n", element->buffer.pointer);
+		{
+			int i, r, at, room_left;
+			const int byte_per_line = 16;
+
+			at = 0;
+			room_left = PAGE_SIZE - 1;
+			for (i = 0; i < element->buffer.length && room_left; i += byte_per_line) {
+				r = hex_dump_to_buffer(element->buffer.pointer + i,
+						       element->buffer.length - i,
+						       byte_per_line, 1, buf + at, room_left,
+						       false);
+				if (r > room_left)
+					goto truncating;
+				at += r;
+				room_left -= r;
+
+				r = sysfs_emit_at(buf, at, "\n");
+				if (!r)
+					goto truncating;
+				at += r;
+				room_left -= r;
+			}
+
+			buf[at] = 0;
+			return at;
+truncating:
+			dev_info(dev, "truncating sysfs content for %s\n", name);
+			sysfs_emit_at(buf, PAGE_SIZE - 4, "..\n");
+			return PAGE_SIZE - 1;
+		}
 	default:
 		dev_err(dev, "element type %d not supported\n", element->type);
 		return -EINVAL;
Signature · 1 line
2.41.0.585.gd2178a4bd4-goog
#1 lore raw
Quoted 6 lines · Tzung-Bi Shih

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

Please add fixes tag and CC stable as well ([email protected]) in next revision.

Fixes 0a4cad9c11ad ("platform/chrome: Add ChromeOS ACPI device driver")
Quoted 52 lines

---
Changes from v1[1]:
- Use hex_dump_to_buffer().
- Rewrite the loop for catching edge cases for truncating.

[1]: https://patchwork.kernel.org/project/chrome-platform/patch/[email protected]/

 drivers/platform/chrome/chromeos_acpi.c | 31 ++++++++++++++++++++++++-
 1 file changed, 30 insertions(+), 1 deletion(-)
drivers/platform/chrome/chromeos_acpi.c +30 -1
index 50d8a4d4352d..761506d307ad 100644
@@ -90,7 +90,36 @@ static int chromeos_acpi_handle_package(struct device *dev, union acpi_object *o
 	case ACPI_TYPE_STRING:
 		return sysfs_emit(buf, "%s\n", element->string.pointer);
 	case ACPI_TYPE_BUFFER:
-		return sysfs_emit(buf, "%s\n", element->buffer.pointer);
+		{
+			int i, r, at, room_left;
+			const int byte_per_line = 16;
+
+			at = 0;
+			room_left = PAGE_SIZE - 1;
+			for (i = 0; i < element->buffer.length && room_left; i += byte_per_line) {
+				r = hex_dump_to_buffer(element->buffer.pointer + i,
+						       element->buffer.length - i,
+						       byte_per_line, 1, buf + at, room_left,
+						       false);
+				if (r > room_left)
+					goto truncating;
+				at += r;
+				room_left -= r;
+
+				r = sysfs_emit_at(buf, at, "\n");
+				if (!r)
+					goto truncating;
+				at += r;
+				room_left -= r;
+			}
+
+			buf[at] = 0;
+			return at;
+truncating:
+			dev_info(dev, "truncating sysfs content for %s\n", name);
+			sysfs_emit_at(buf, PAGE_SIZE - 4, "..\n");
+			return PAGE_SIZE - 1;
+		}
 	default:
 		dev_err(dev, "element type %d not supported\n", element->type);
 		return -EINVAL;
Signature · 2 lines
BR,
Muhammad Usama Anjum
author maintainer
#2 lore raw
Quoted 11 lines · Muhammad Usama Anjum
Quoted 6 lines · Tzung-Bi Shih

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

Please add fixes tag and CC stable as well ([email protected]) in next revision.

Fixes 0a4cad9c11ad ("platform/chrome: Add ChromeOS ACPI device driver")

I think a Fixes tag should be sufficient.

Fix in v3 (https://patchwork.kernel.org/project/chrome-platform/patch/[email protected]/).

#3 lore raw
Quoted 7 lines · Tzung-Bi Shih
Quoted 6 lines · Muhammad Usama Anjum
Quoted 5 lines · Tzung-Bi Shih

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

Please add fixes tag and CC stable as well ([email protected]) in next revision.

Fixes 0a4cad9c11ad ("platform/chrome: Add ChromeOS ACPI device driver")

I think a Fixes tag should be sufficient.

Why do you think so? As this is a bug fix which exists in the driver since it was written, if we add fixes tag and probably cc stable as well, the patch will be applied to all LTS kernels which we really want. In this way, the bug would be fixed on all LTS kernels not just the latest kernel.

Quoted 2 lines

Fix in v3 (https://patchwork.kernel.org/project/chrome-platform/patch/[email protected]/).

Signature · 2 lines
BR,
Muhammad Usama Anjum
author maintainer
#4 lore raw
Quoted 12 lines · Muhammad Usama Anjum
Quoted 11 lines · Tzung-Bi Shih
Quoted 10 lines · Muhammad Usama Anjum
Quoted 6 lines · Tzung-Bi Shih

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

Please add fixes tag and CC stable as well ([email protected]) in next revision.

Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Fixes 0a4cad9c11ad ("platform/chrome: Add ChromeOS ACPI device driver")

I think a Fixes tag should be sufficient.

Why do you think so? As this is a bug fix which exists in the driver since it was written, if we add fixes tag and probably cc stable as well, the patch will be applied to all LTS kernels which we really want. In this way, the bug would be fixed on all LTS kernels not just the latest kernel.

Without the patch, I observed some malfunctions of ChromiumOS userland program crossystem. There is no kernel crashes and the system is bootable. Thus, I think a Fixes tag should be sufficient.

author maintainer
#5 lore raw
Quoted 16 lines · Tzung-Bi Shih
Quoted 15 lines · Muhammad Usama Anjum
Quoted 14 lines · Tzung-Bi Shih
Quoted 11 lines · Muhammad Usama Anjum
Quoted 6 lines · Tzung-Bi Shih

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

Please add fixes tag and CC stable as well ([email protected]) in next revision.

Fixes 0a4cad9c11ad ("platform/chrome: Add ChromeOS ACPI device driver")

I think a Fixes tag should be sufficient.

Quoted 8 lines
Quoted 4 lines

Why do you think so? As this is a bug fix which exists in the driver since it was written, if we add fixes tag and probably cc stable as well, the patch will be applied to all LTS kernels which we really want. In this way, the bug would be fixed on all LTS kernels not just the latest kernel.

Without the patch, I observed some malfunctions of ChromiumOS userland program crossystem. There is no kernel crashes and the system is bootable. Thus, I think a Fixes tag should be sufficient.

Let's cc stable as well.

#6 lore raw
In reply to Tzung-Bi Shih
Quoted 43 lines · Tzung-Bi Shih

element->buffer.pointer should be binary blob. %s doesn't work perfect for them.

Print hex string for ACPI_TYPE_BUFFER.

---
Changes from v1[1]:
- Use hex_dump_to_buffer().
- Rewrite the loop for catching edge cases for truncating.

[1]: https://patchwork.kernel.org/project/chrome-platform/patch/[email protected]/

 drivers/platform/chrome/chromeos_acpi.c | 31 ++++++++++++++++++++++++-
 1 file changed, 30 insertions(+), 1 deletion(-)
drivers/platform/chrome/chromeos_acpi.c +17 -1
index 50d8a4d4352d..761506d307ad 100644
@@ -90,7 +90,36 @@ static int chromeos_acpi_handle_package(struct device *dev, union acpi_object *o
        case ACPI_TYPE_STRING:
                return sysfs_emit(buf, "%s\n", element->string.pointer);
        case ACPI_TYPE_BUFFER:
-               return sysfs_emit(buf, "%s\n", element->buffer.pointer);
+               {
+                       int i, r, at, room_left;
+                       const int byte_per_line = 16;
+
+                       at = 0;
+                       room_left = PAGE_SIZE - 1;
+                       for (i = 0; i < element->buffer.length && room_left; i += byte_per_line) {
+                               r = hex_dump_to_buffer(element->buffer.pointer + i,
+                                                      element->buffer.length - i,
+                                                      byte_per_line, 1, buf + at, room_left,
+                                                      false);
+                               if (r > room_left)
+                                       goto truncating;
+                               at += r;
+                               room_left -= r;
+
+                               r = sysfs_emit_at(buf, at, "\n");

drivers/platform/chrome/chromeos_acpi.c static int chromeos_acpi_handle_package(struct device *dev, union acpi_object *o · line 119

Quoted 10 lines
drivers/platform/chrome/chromeos_acpi.c +10 -0
+                               if (!r)
+                                       goto truncating;
+                               at += r;
+                               room_left -= r;
+                       }
+
+                       buf[at] = 0;
+                       return at;
+truncating:
+                       dev_info(dev, "truncating sysfs content for %s\n", name);

I really dislike that because it will, if it happens, clog the log (the condition will be permanent). How about using dev_info_once() ?

Thanks,
Guenter

Quoted 9 lines
drivers/platform/chrome/chromeos_acpi.c +3 -1
+                       sysfs_emit_at(buf, PAGE_SIZE - 4, "..\n");
+                       return PAGE_SIZE - 1;
+               }
        default:
                dev_err(dev, "element type %d not supported\n", element->type);
                return -EINVAL;
--

2.41.0.585.gd2178a4bd4-goog

author maintainer
#7 lore raw
Quoted 6 lines · Guenter Roeck

drivers/platform/chrome/chromeos_acpi.c static int chromeos_acpi_handle_package(struct device *dev, union acpi_object *o · line 119

Quoted 2 lines · Tzung-Bi Shih
drivers/platform/chrome/chromeos_acpi.c +2 -0
+truncating:
+                       dev_info(dev, "truncating sysfs content for %s\n", name);

I really dislike that because it will, if it happens, clog the log (the condition will be permanent). How about using dev_info_once() ?

Fix in v3(https://patchwork.kernel.org/project/chrome-platform/patch/[email protected]/).