Re: [PATCH v5 2/7] firmware: add offset to request_firmware_into_buf
From: Luis Chamberlain
Date: Tue May 12 2020 - 20:33:07 EST
On Thu, May 07, 2020 at 05:27:34PM -0700, Scott Branden wrote:
> Add offset to request_firmware_into_buf to allow for portions
> of firmware file to be read into a buffer. Necessary where firmware
> needs to be loaded in portions from file in memory constrained systems.
>
> Signed-off-by: Scott Branden <scott.branden@xxxxxxxxxxxx>
> ---
> drivers/base/firmware_loader/firmware.h | 5 +++
> drivers/base/firmware_loader/main.c | 52 +++++++++++++++++--------
> drivers/soc/qcom/mdt_loader.c | 7 +++-
> include/linux/firmware.h | 8 +++-
> lib/test_firmware.c | 4 +-
> 5 files changed, 55 insertions(+), 21 deletions(-)
>
> diff --git a/drivers/base/firmware_loader/firmware.h b/drivers/base/firmware_loader/firmware.h
> index 25836a6afc9f..1147dae01148 100644
> --- a/drivers/base/firmware_loader/firmware.h
> +++ b/drivers/base/firmware_loader/firmware.h
> @@ -32,6 +32,8 @@
> * @FW_OPT_FALLBACK_PLATFORM: Enable fallback to device fw copy embedded in
> * the platform's main firmware. If both this fallback and the sysfs
> * fallback are enabled, then this fallback will be tried first.
> + * @FW_OPT_PARTIAL: Allow partial read of firmware instead of needing to read
> + * entire file.
See, this allows us use kdoc to document his nicely. Do that with the
kernel pread stuff.
> @@ -68,6 +71,8 @@ struct fw_priv {
> void *data;
> size_t size;
> size_t allocated_size;
> + size_t offset;
> + unsigned int flags;
But flags is a misnomer, you just do two operations, just juse an enum
here to classify the read operation.
> index 76f79913916d..4552b7bb819f 100644
> --- a/drivers/base/firmware_loader/main.c
> +++ b/drivers/base/firmware_loader/main.c
> @@ -167,7 +167,8 @@ static int fw_cache_piggyback_on_request(const char *name);
>
> static struct fw_priv *__allocate_fw_priv(const char *fw_name,
> struct firmware_cache *fwc,
> - void *dbuf, size_t size)
> + void *dbuf, size_t size,
> + size_t offset, unsigned int flags)
And likewise just use an enum here too.
> @@ -210,9 +213,11 @@ static struct fw_priv *__lookup_fw_priv(const char *fw_name)
> static int alloc_lookup_fw_priv(const char *fw_name,
> struct firmware_cache *fwc,
> struct fw_priv **fw_priv, void *dbuf,
> - size_t size, enum fw_opt opt_flags)
> + size_t size, enum fw_opt opt_flags,
> + size_t offset)
flags? But its a single variable enum!
> {
> struct fw_priv *tmp;
> + unsigned int pread_flags;
>
> spin_lock(&fwc->lock);
> if (!(opt_flags & FW_OPT_NOCACHE)) {
> @@ -226,7 +231,12 @@ static int alloc_lookup_fw_priv(const char *fw_name,
> }
> }
>
> - tmp = __allocate_fw_priv(fw_name, fwc, dbuf, size);
> + if (opt_flags & FW_OPT_PARTIAL)
> + pread_flags = KERNEL_PREAD_FLAG_PART;
> + else
> + pread_flags = KERNEL_PREAD_FLAG_WHOLE;
> +
> + tmp = __allocate_fw_priv(fw_name, fwc, dbuf, size, offset, pread_flags);
One of the advantages of using an enum is that you can then use a switch
here, and the compiler will warn if you haven't handled all the cases.
> /* load firmware files from the mount namespace of init */
> - rc = kernel_read_file_from_path_initns(path, &buffer,
> - &size, msize, id);
> + rc = kernel_pread_file_from_path_initns(path, &buffer,
> + &size, fw_priv->offset,
> + msize,
> + fw_priv->flags, id);
And here you'd just pass the kernel enum.
You get the idea, I stopped reviewing after this.
Luis