Hello i2c devs,
This code generatates the following static checker warning:
drivers/i2c/i2c-core-acpi.c:615 i2c_acpi_space_handler()
warn: assigning signed to unsigned: 'gsb->status = status' 's32min-s32max'
drivers/i2c/i2c-core-acpi.c
584 case ACPI_GSB_ACCESS_ATTRIB_BLOCK:
585 if (action == ACPI_READ) {
586 status = i2c_smbus_read_block_data(client, command,
587 gsb->data);
588 if (status >= 0) {
589 gsb->len = status;
590 status = 0;
591 }
592 } else {
593 status = i2c_smbus_write_block_data(client, command,
594 gsb->len, gsb->data);
595 }
596 break;
597
598 case ACPI_GSB_ACCESS_ATTRIB_MULTIBYTE:
599 if (action == ACPI_READ) {
600 status = acpi_gsb_i2c_read_bytes(client, command,
601 gsb->data, info->access_length);
602 } else {
603 status = acpi_gsb_i2c_write_bytes(client, command,
604 gsb->data, info->access_length);
605 }
606 break;
607
608 default:
609 dev_warn(&adapter->dev, "protocol 0x%02x not supported for client 0x%02x\n",
610 accessor_type, client->addr);
611 ret = AE_BAD_PARAMETER;
612 goto err;
613 }
614
615 gsb->status = status;
^^^^^^^^^^^^^^^^^^^^
Status can be a negative error code, and we're saving it to a u8. I
think returning success is intended on this path, so that's probably
fine.
616
617 err:
618 kfree(client);
619 ACPI_FREE(ares);
620 return ret;
621 }
regards,
dan carpenter
