| 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
| ||
|
Date: Thu, 25 Jan 2018 10:30:57 +0100
From: Greg KH <gregkh@...uxfoundation.org>
To: Jolly Shah <jolly.shah@...inx.com>
Cc: ard.biesheuvel@...aro.org, mingo@...nel.org,
matt@...eblueprint.co.uk, sudeep.holla@....com,
hkallweit1@...il.com, keescook@...omium.org,
dmitry.torokhov@...il.com, michal.simek@...inx.com,
robh+dt@...nel.org, mark.rutland@....com,
linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
devicetree@...r.kernel.org, Jolly Shah <jollys@...inx.com>,
Rajan Vaja <rajanv@...inx.com>
Subject: Re: [PATCH v3 4/4] drivers: firmware: xilinx: Add debugfs interface
On Wed, Jan 24, 2018 at 03:21:14PM -0800, Jolly Shah wrote:
> Firmware-debug provides debugfs interface to all APIs.
I don't understand this changelog comment, care to make it more
informational? At least describe the debugfs files you are adding so
that people have a chance to understand what is going on here :)
> diff --git a/drivers/firmware/xilinx/zynqmp/Kconfig b/drivers/firmware/xilinx/zynqmp/Kconfig
> index 8f7709d..bdd0188 100644
> --- a/drivers/firmware/xilinx/zynqmp/Kconfig
> +++ b/drivers/firmware/xilinx/zynqmp/Kconfig
> @@ -13,4 +13,11 @@ config ZYNQMP_FIRMWARE
> Say yes to enable zynqmp firmware interface driver.
> In doubt, say N
>
> +config ZYNQMP_FIRMWARE_DEBUG
> + bool "Enable Xilinx Zynq MPSoC firmware debug APIs"
> + default ARCH_ZYNQMP && ZYNQMP_FIRMWARE && DEBUG_FS
> + help
> + Say yes to enable zynqmp firmware interface debug APIs.
> + In doubt, say N
So your default is going to be Y if the driver is selected? That's not
good, just leave the default alone please.
> +
> endmenu
> diff --git a/drivers/firmware/xilinx/zynqmp/Makefile b/drivers/firmware/xilinx/zynqmp/Makefile
> index 6629781..02f0c9a 100644
> --- a/drivers/firmware/xilinx/zynqmp/Makefile
> +++ b/drivers/firmware/xilinx/zynqmp/Makefile
> @@ -2,3 +2,4 @@
> # Makefile for Xilinx firmwares
>
> obj-$(CONFIG_ZYNQMP_FIRMWARE) += firmware.o firmware-ggs.o
> +obj-$(CONFIG_ZYNQMP_FIRMWARE_DEBUG) += firmware-debug.o
> diff --git a/drivers/firmware/xilinx/zynqmp/firmware-debug.c b/drivers/firmware/xilinx/zynqmp/firmware-debug.c
> new file mode 100644
> index 0000000..daefc62
> --- /dev/null
> +++ b/drivers/firmware/xilinx/zynqmp/firmware-debug.c
> @@ -0,0 +1,511 @@
> +// SPDX-License-Identifier: GPL-2.0+
> +/*
> + * Xilinx Zynq MPSoC Firmware layer for debugfs APIs
> + *
> + * Copyright (C) 2014-2018 Xilinx, Inc.
> + *
> + * Michal Simek <michal.simek@...inx.com>
> + * Davorin Mista <davorin.mista@...ios.com>
> + * Jolly Shah <jollys@...inx.com>
> + * Rajan Vaja <rajanv@...inx.com>
> + */
> +
> +#include <linux/compiler.h>
> +#include <linux/module.h>
> +#include <linux/slab.h>
> +#include <linux/debugfs.h>
> +#include <linux/uaccess.h>
> +#include <linux/firmware/xilinx/zynqmp/firmware.h>
> +#include <linux/firmware/xilinx/zynqmp/firmware-debug.h>
> +
> +#define DRIVER_NAME "zynqmp-firmware"
You define this in 2 places, but only use it in one, you don't need this
at all, please remove all instances.
> +static int process_api_request(u32 pm_id, u64 *pm_api_arg, u32 *pm_api_ret)
> +{
> + const struct zynqmp_eemi_ops *eemi_ops = get_eemi_ops();
> + u32 pm_api_version;
> + int ret;
> +
> + if (!eemi_ops)
> + return -ENXIO;
> +
> + switch (pm_id) {
> + case PM_GET_API_VERSION:
> + eemi_ops->get_api_version(&pm_api_version);
> + pr_info("%s PM-API Version = %d.%d\n", __func__,
> + pm_api_version >> 16, pm_api_version & 0xffff);
This is a _very_ noisy function, dumping a lot of stuff to the kernel
log. Why? Why not send it back to the caller of the debugfs file
reader instead?
> + case PM_GET_NODE_STATUS:
> + ret = eemi_ops->get_node_status(pm_api_arg[0],
> + &pm_api_ret[0],
> + &pm_api_ret[1],
> + &pm_api_ret[2]);
> + if (!ret)
> + pr_info("GET_NODE_STATUS:\n\tNodeId: %llu\n\tStatus: %u\n\tRequirements: %u\n\tUsage: %u\n",
> + pm_api_arg[0], pm_api_ret[0],
> + pm_api_ret[1], pm_api_ret[2]);
Multi-line dmesg messages? Not good, you just lost the logging level
for the multiple lines :(
Again, don't do this at all, just put it in file output instead. That
way tools/users can actually use it, instead of having to dig through
kernel log messages.
> +/**
> + * zynqmp_pm_api_debugfs_init - Initialize debugfs interface
> + *
> + * Return: None
> + */
> +void zynqmp_pm_api_debugfs_init(void)
> +{
> + struct dentry *root_dir;
> + struct dentry *d;
> +
> + /* Initialize debugfs interface */
> + root_dir = debugfs_create_dir(DRIVER_NAME, NULL);
> + if (!root_dir) {
> + pr_warn("debugfs_create_dir failed\n");
Why warn? What can a user do about this? Just return and move on.
> + return;
> + }
> +
> + d = debugfs_create_file("pm", 0220, root_dir, NULL,
> + &fops_zynqmp_pm_dbgfs);
> + if (!d) {
> + pr_warn("debugfs_create_file power failed\n");
> + goto err_dbgfs;
> + }
> +
> + d = debugfs_create_file("api_version", 0444, root_dir,
> + NULL, &fops_zynqmp_pm_dbgfs);
> + if (!d) {
> + pr_warn("debugfs_create_file api_version failed\n");
> + goto err_dbgfs;
Same for these files, who cares if they were not created or not at all?
No need to even check here at all, this whole function can be reduced to just 3 lines:
debugfs_create_dir(...);
debugfs_create_file(...);
debugfs_create_file(...);
There, nice and simple. Remember, debugfs doesn't matter, and it should
be _really_ easy to use. Don't make it more complex than it has to be
please.
thanks,
greg k-h
Powered by blists - more mailing lists