Differential D27414 Diff 83055 sys/compat/linuxkpi/common/src/linux_firmware.c

Changeset View

Standalone View

sys/compat/linuxkpi/common/src/linux_firmware.c

This file was added.

				/*-
				* SPDX-License-Identifier: BSD-2-Clause
				*
				* Copyright (c) 2020-2021 The FreeBSD Foundation
				*
				* This software was developed by Björn Zeeb under sponsorship from
				* the FreeBSD Foundation.
				*
				* Redistribution and use in source and binary forms, with or without
				* modification, are permitted provided that the following conditions
				* are met:
				* 1. Redistributions of source code must retain the above copyright
				* notice, this list of conditions and the following disclaimer.
				* 2. Redistributions in binary form must reproduce the above copyright
				* notice, this list of conditions and the following disclaimer in the
				* documentation and/or other materials provided with the distribution.
				*
				* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
				* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
				* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
				* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
				* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
				* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
				* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
				* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
				* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
				* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
				* SUCH DAMAGE.
				*
				* $FreeBSD$
				*/

				#include <sys/types.h>
				#include <sys/malloc.h>
				#include <sys/firmware.h>

				#include <linux/types.h>
				#include <linux/device.h>

				#include <linux/firmware.h>
				#undef firmware

				MALLOC_DEFINE(M_LKPI_FW, "lkpifw", "LinuxKPI firmware");

				static int
				_linuxkpi_request_firmware(const char fw_name, const struct linuxkpi_firmware *fw,
				struct device *dev, gfp_t gfp __unused, bool enoentok, bool warn)
				{
				const struct firmware *fbdfw;
				struct linuxkpi_firmware *lfw;
				const char *fwimg;
				char *p;
				uint32_t flags;

				if (fw_name == NULL \|\| fw == NULL \|\| dev == NULL)
				return (-EINVAL);

				/* Set independent on "warn". To debug, bootverbose is avail. */
				flags = FIRMWARE_GET_NOWARN;

				KASSERT(gfp == GFP_KERNEL, ("%s: gfp %#x\n", __func__, gfp));
				lfw = malloc(sizeof(*lfw), M_LKPI_FW, M_WAITOK \| M_ZERO);

				/*
				* Linux can have a path in the firmware which is hard to replicate
				* for auto-firmware-module-loading.
				* On FreeBSD, depending on what people do, the firmware will either
				* be called "fw", or "dir_fw", or "modname_dir_fw". The latter the
				* driver author has to deal with herself (requesting the special name).
				* We also optionally flatten '/'s and '.'s as some firmware modules do.
				* We probe in the least-of-work order avoiding memory operations.
				* It will be preferred to build the firmware .ko in a well matching
				* way rather than adding more name-mangling-hacks here in the future
				* (though we could if needed).
				*/
				/* (1) Try any name removed of path. */
				fwimg = strrchr(fw_name, '/');
				if (fwimg != NULL)
				fwimg++;
				if (fwimg == NULL \|\| *fwimg == '\0')
				fwimg = fw_name;
				fbdfw = firmware_get_flags(fwimg, flags);
				/* (2) Try the original name if we have not yet. */
				if (fbdfw == NULL && fwimg != fw_name) {
				fwimg = fw_name;
				fbdfw = firmware_get_flags(fwimg, flags);
				}
				/* (3) Flatten '/' and then '.' to '_' and try with adjusted name. */
				if (fbdfw == NULL &&
				(strchr(fw_name, '/') != NULL \|\| strchr(fw_name, '.') != NULL)) {
				fwimg = strdup(fw_name, M_LKPI_FW);
				if (fwimg != NULL) {
				while ((p = strchr(fwimg, '/')) != NULL)
				*p = '_';
				fbdfw = firmware_get_flags(fwimg, flags);
				if (fbdfw == NULL) {
				while ((p = strchr(fwimg, '.')) != NULL)
				*p = '_';
				fbdfw = firmware_get_flags(fwimg, flags);
				}
				free(__DECONST(void *, fwimg), M_LKPI_FW);
				}
				}
				if (fbdfw == NULL) {
				if (enoentok)
				*fw = lfw;
				else {
				free(lfw, M_LKPI_FW);
				*fw = NULL;
				}
				if (warn)
				device_printf(dev->bsddev, "could not load firmware "
				"image '%s'\n", fw_name);
				manuUnsubmitted Done Inline Actions Using the bsddev and device_printf would be better. manu: Using the bsddev and device_printf would be better.
				bzAuthorUnsubmitted Done Inline Actions Done. Thanks. Good idea! bz: Done. Thanks. Good idea!
				return (-ENOENT);
				}

				device_printf(dev->bsddev,"successfully loaded firmware image '%s'\n",
				fw_name);
				manuUnsubmitted Done Inline Actions Same here. manu: Same here.
				lfw->fbdfw = fbdfw;
				lfw->data = (const uint8_t *)fbdfw->data;
				lfw->size = fbdfw->datasize;
				*fw = lfw;
				return (0);
				}

				int
				linuxkpi_request_firmware_nowait(struct module *mod __unused, bool _t __unused,
				const char fw_name, struct device dev, gfp_t gfp, void *drv,
				void(cont)(const struct linuxkpi_firmware , void *))
				{
				const struct linuxkpi_firmware *lfw;
				int error;

				/*
				* Linux seems to run the callback if it cannot find the firmware.
				* The fact that this is "_nowait()" and has a callback seems to
				* imply that this is run in a deferred conext which we currently
				* do not do. Should it become necessary (a driver actually requiring
				manuUnsubmitted Done Inline Actions Is it something that Linux does ? (Running the callback defered). If yes is it something that you plan to add support to ? manu: Is it something that Linux does ? (Running the callback defered). If yes is it something that…
				bzAuthorUnsubmitted Done Inline Actions I currently do not plan to implement it. None of the drivers I have seen so far and tested do require it. I updated the comment to be more clear. The comment is merely left that in case in 10 years someone works on a driver which needs the deferred context they can find the problem more easily. Should I come across a driver needing it, I will of course implement it. bz: I currently do not plan to implement it. None of the drivers I have seen so far and tested do…
				* it) we would need to implement it here.
				*/
				error = _linuxkpi_request_firmware(fw_name, &lfw, dev, gfp, true, true);
				if (error == -ENOENT)
				error = 0;
				if (error == 0)
				cont(lfw, drv);

				return (error);
				}

				int
				linuxkpi_request_firmware(const struct linuxkpi_firmware **fw,
				const char fw_name, struct device dev)
				{

				return (_linuxkpi_request_firmware(fw_name, fw, dev, GFP_KERNEL, false,
				true));
				}

				int
				linuxkpi_firmware_request_nowarn(const struct linuxkpi_firmware **fw,
				const char fw_name, struct device dev)
				{

				return (_linuxkpi_request_firmware(fw_name, fw, dev, GFP_KERNEL, false,
				false));
				}

				void
				linuxkpi_release_firmware(const struct linuxkpi_firmware *fw)
				{

				if (fw == NULL)
				return;

				if (fw->fbdfw)
				firmware_put(fw->fbdfw, FIRMWARE_UNLOAD);
				free(__DECONST(void *, fw), M_LKPI_FW);
				}