From: AKASHI Takahiro <takahiro.akashi@linaro.org>
To: Masahisa Kojima <masahisa.kojima@linaro.org>
Cc: u-boot@lists.denx.de, Heinrich Schuchardt <xypron.glpk@gmx.de>,
Ilias Apalodimas <ilias.apalodimas@linaro.org>
Subject: Re: [PATCH 2/2] efi_loader: support boot from URI device path
Date: Thu, 24 Aug 2023 11:24:31 +0900 [thread overview]
Message-ID: <ZOa/X1DMZOMgotSB@octopus> (raw)
In-Reply-To: <20230823083721.1318097-3-masahisa.kojima@linaro.org>
Kojima-san,
On Wed, Aug 23, 2023 at 05:37:20PM +0900, Masahisa Kojima wrote:
> This supports to boot from the URI device path.
> When user selects the URI device path, bootmgr downloads
> the file using wget into the address specified by loadaddr
> env variable.
> If the file is .iso or .img file, mount the image with blkmap
> then try to boot with the default file(e.g. EFI/BOOT/BOOTAA64.EFI).
> If the file is .efi file, load and start the downloaded file.
Is this behavior part of UEFI specification?
Even so, it would be better to describe it in uefi.rst (or else),
including URI usage.
> Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
> ---
> lib/efi_loader/efi_bootmgr.c | 213 +++++++++++++++++++++++++++++++++++
> 1 file changed, 213 insertions(+)
>
> diff --git a/lib/efi_loader/efi_bootmgr.c b/lib/efi_loader/efi_bootmgr.c
> index a40762c74c..8b20f486f2 100644
> --- a/lib/efi_loader/efi_bootmgr.c
> +++ b/lib/efi_loader/efi_bootmgr.c
> @@ -7,10 +7,14 @@
>
> #define LOG_CATEGORY LOGC_EFI
>
> +#include <blk.h>
> +#include <blkmap.h>
> #include <common.h>
> #include <charset.h>
> +#include <dm.h>
> #include <log.h>
> #include <malloc.h>
> +#include <net.h>
> #include <efi_default_filename.h>
> #include <efi_loader.h>
> #include <efi_variable.h>
> @@ -168,6 +172,209 @@ out:
> return ret;
> }
>
> +#if (IS_ENABLED(CONFIG_BLKMAP) && IS_ENABLED(CONFIG_CMD_WGET) && IS_ENABLED(CONFIG_CMD_DNS))
> +/**
> + * mount_image() - mount the image
> + *
> + * @lo_label label of load option
> + * @file_size file size
> + * @handle: pointer to handle for newly installed image
> + * Return: status code
> + */
> +static efi_status_t mount_image(u16 *lo_label, int file_size,
> + efi_handle_t *handle)
I wonder why not adding "address" parameter to make this function
more generic as Simon suggested.
> +{
> + int err;
> + efi_status_t ret;
> + char *label = NULL, *p;
> + lbaint_t blknum;
> + struct udevice *bm_dev;
> + efi_handle_t bm_handle;
> + struct udevice *blk, *partition;
> + struct efi_handler *handler;
> + struct efi_device_path *file_path;
> + struct efi_device_path *device_path;
> +
> + label = efi_alloc(utf16_utf8_strlen(lo_label) + 1);
> + if (!label)
> + return EFI_OUT_OF_RESOURCES;
> +
> + p = label;
> + utf16_utf8_strcpy(&p, lo_label);
> + err = blkmap_create(label, NULL);
> + if (err) {
> + log_err("failed to create blkmap\n");
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> + bm_dev = blkmap_from_label(label);
> + if (!bm_dev) {
> + log_err("\"%s\" is not the name of any known blkmap\n", label);
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> +
> + blknum = file_size / 512; /* TODO: don't use literal value. */
> + err = blkmap_map_pmem(bm_dev, 0, blknum, image_load_addr);
> + if (err) {
> + log_err("Unable to map %#llx at block %d : %d\n",
> + (unsigned long long)image_load_addr, 0, err);
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> + log_info("Block %d+0x" LBAF " mapped to %#llx\n", 0, blknum,
> + (unsigned long long)image_load_addr);
> +
> + /* TODO: without calling this, partition devices are not binded. */
> + blk_list_part(UCLASS_BLKMAP);
I think that blkmap should provide a way to *probe* its child block
device. In fact,
struct blkmap *bm = dev_get_plat(bm_dev);
device_probe(bm->blk);
should work. (Not tested though)
> +
> + /*
> + * Search the partition having EFI_SIMPLE_FILE_SYSTEM_PROTOCOL,
> + * then try to load with the default boot file(e.g. EFI/BOOT/BOOTAA64.EFI).
> + */
> + device_foreach_child(blk, bm_dev)
Does bm_dev have more than one block devices?
> + {
> + device_foreach_child(partition, blk)
> + {
> + if (dev_tag_get_ptr(partition, DM_TAG_EFI,
> + (void **)&bm_handle)) {
> + log_warning("DM_TAG_EFI not found\n");
> + continue;
> + }
> +
> + ret = efi_search_protocol(
> + bm_handle,
> + &efi_simple_file_system_protocol_guid,
> + &handler);
> + if (ret != EFI_SUCCESS)
> + continue;
> +
> + ret = efi_search_protocol(
> + bm_handle, &efi_guid_device_path, &handler);
> + if (ret != EFI_SUCCESS)
> + continue;
> +
> + ret = efi_protocol_open(handler, (void **)&device_path,
> + efi_root, NULL,
> + EFI_OPEN_PROTOCOL_GET_PROTOCOL);
> + if (ret != EFI_SUCCESS)
> + continue;
> +
> + file_path = expand_media_path(device_path);
> + ret = EFI_CALL(efi_load_image(true, efi_root, file_path,
> + NULL, 0, handle));
> + efi_free_pool(file_path);
> + if (ret == EFI_SUCCESS)
> + goto out;
> + }
> + }
Another idea would be to create a new boot option (as a removable
media) and to invoke it from bootmgr.
> +
> +out:
> + efi_free_pool(label);
> +
> + return ret;
> +}
> +
> +/**
> + * try_load_from_uri_path() - Handle the URI device path
> + *
> + * @uridp: uri device path
> + * @lo_label label of load option
> + * @handle: pointer to handle for newly installed image
> + * Return: status code
> + */
> +static efi_status_t try_load_from_uri_path(struct efi_device_path_uri *uridp,
> + u16 *lo_label,
> + efi_handle_t *handle)
> +{
> + efi_status_t ret;
> + int file_size, file_name_len;
> + char *s, *host_name, *file_name, *str_copy;
> +
> + /*
> + * Download file using wget.
> + *
> + * URI device path content is like http://www.example.com/sample/test.iso.
> + * U-Boot wget takes the target uri in this format.
> + * "<http server ip>:<file path>" e.g.) 192.168.1.1:/sample/test.iso
> + * Need to resolve the http server ip address before starting wget.
> + */
> +
> + /* only support "http://" */
> + if (strncmp(uridp->uri, "http://", 7)) {
> + log_err("Error: uri must start with http://\n");
> + return EFI_INVALID_PARAMETER;
> + }
> +
> + str_copy = strdup(uridp->uri);
> + if (!str_copy)
> + return EFI_OUT_OF_RESOURCES;
> + s = str_copy + strlen("http://");
> + host_name = strsep(&s, "/");
> + if (!s) {
> + log_err("Error: invalied uri, no file path\n");
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> + file_name = s;
> + net_dns_resolve = host_name;
> + net_dns_env_var = "httpserverip";
> + if (net_loop(DNS) < 0) {
> + log_err("Error: dns lookup of %s failed, check setup\n", net_dns_resolve);
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> + s = env_get("httpserverip");
> + if (!s) {
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> +
> + /*
> + * WGET requires that "net_boot_file_name" and "image_load_addr" global
> + * variables are properly set in advance.
> + */
> + strlcpy(net_boot_file_name, s, 1024);
> + strlcat(net_boot_file_name, ":/", 1024); /* append '/' which is removed by strsep() */
> + strlcat(net_boot_file_name, file_name, 1024);
> + s = env_get("loadaddr");
> + if (!s) {
> + log_err("Error: loadaddr is not set\n");
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> + image_load_addr = hextoul(s, NULL);
> +
> + file_size = net_loop(WGET);
> + if (file_size < 0) {
> + log_err("Error: downloading file failed\n");
> + ret = EFI_INVALID_PARAMETER;
> + goto out;
> + }
> +
> + /*
> + * Identify file type by file extension.
> + * If the file extension is ".iso" or ".img", mount it and boot with default file.
> + * If the file is ".efi", load and start the downloaded file.
> + */
> + file_name_len = strlen(net_boot_file_name);
> + if (!strncmp(&net_boot_file_name[file_name_len - 4], ".iso", 4) ||
> + !strncmp(&net_boot_file_name[file_name_len - 4], ".img", 4)) {
> + ret = mount_image(lo_label, file_size, handle);
> + } else if (!strncmp(&net_boot_file_name[file_name_len - 4], ".efi", 4)) {
> + ret = efi_run_image((void *)image_load_addr, file_size);
Nak.
try_load_from_uri_path() is called from try_load_entry().
This function's purpose is solely to load an image.
The execution should be triggered later.
-Takahiro Akashi
> + } else {
> + log_err("Error: file type is not supported\n");
> + ret = EFI_INVALID_PARAMETER;
> + }
> +
> +out:
> + free(str_copy);
> +
> + return ret;
> +}
> +#endif
> +
> /**
> * try_load_entry() - try to load image for boot option
> *
> @@ -211,6 +418,12 @@ static efi_status_t try_load_entry(u16 n, efi_handle_t *handle,
> if (EFI_DP_TYPE(lo.file_path, MEDIA_DEVICE, FILE_PATH)) {
> /* file_path doesn't contain a device path */
> ret = try_load_from_short_path(lo.file_path, handle);
> +#if (IS_ENABLED(CONFIG_BLKMAP) && IS_ENABLED(CONFIG_CMD_WGET) && IS_ENABLED(CONFIG_CMD_DNS))
> + } else if (EFI_DP_TYPE(lo.file_path, MESSAGING_DEVICE, MSG_URI)) {
> + ret = try_load_from_uri_path(
> + (struct efi_device_path_uri *)lo.file_path,
> + lo.label, handle);
> +#endif
> } else {
> file_path = expand_media_path(lo.file_path);
> ret = EFI_CALL(efi_load_image(true, efi_root, file_path,
> --
> 2.34.1
>
next prev parent reply other threads:[~2023-08-24 2:24 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-23 8:37 [PATCH 0/2] Add EFI HTTP boot support Masahisa Kojima
2023-08-23 8:37 ` [PATCH 1/2] cmd: efidebug: add uri device path Masahisa Kojima
2023-08-24 0:11 ` AKASHI Takahiro
2023-08-25 5:57 ` Masahisa Kojima
2023-08-26 2:21 ` Heinrich Schuchardt
2023-08-28 6:02 ` Masahisa Kojima
2023-08-24 5:38 ` Heinrich Schuchardt
2023-08-25 5:58 ` Masahisa Kojima
2023-08-23 8:37 ` [PATCH 2/2] efi_loader: support boot from URI " Masahisa Kojima
2023-08-23 23:57 ` Simon Glass
2023-08-25 6:31 ` Masahisa Kojima
2023-08-24 2:24 ` AKASHI Takahiro [this message]
2023-08-24 2:32 ` AKASHI Takahiro
2023-08-25 6:40 ` Masahisa Kojima
2023-08-24 6:58 ` Heinrich Schuchardt
2023-08-25 7:56 ` Masahisa Kojima
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZOa/X1DMZOMgotSB@octopus \
--to=takahiro.akashi@linaro.org \
--cc=ilias.apalodimas@linaro.org \
--cc=masahisa.kojima@linaro.org \
--cc=u-boot@lists.denx.de \
--cc=xypron.glpk@gmx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox