From: Michael Olbrich <m.olbrich@pengutronix.de>
To: Roland Hieber <rhi@pengutronix.de>
Cc: distrokit@pengutronix.de
Subject: Re: [DistroKit] [PATCH 5/6] images: remove qemu size hack
Date: Wed, 11 Oct 2023 08:51:15 +0200 [thread overview]
Message-ID: <ZSZF43fZNs8TehiL@pengutronix.de> (raw)
In-Reply-To: <20231010191631.31141-5-rhi@pengutronix.de>
On Tue, Oct 10, 2023 at 09:16:30PM +0200, Roland Hieber wrote:
> genimage knows about the 'size' option for images to fill the image to a
> certain size, and since genimage v14 it supports the 'fill' option to
> write all zeroes at the end of the image and not cut them off to
> optimise the image size. Use this to blow up the images that are used
> with qemu to the next power of two so qemu can work with them, and
> remove our invisible-partition-at-the-end hack.
>
> On MIPS, we also need to decrease the size of the last partition so that
> it still fits inside the specified image size, because the first MiB
> (when using 1 MiB alignment) is occupied by the partition table, so the
> first partition only starts at offset 1 MiB.
>
> Signed-off-by: Roland Hieber <rhi@pengutronix.de>
> ---
> .../platform-mips/config/images/ar9331.config | 16 ++++------------
> configs/platform-mips/config/images/malta.config | 16 ++++------------
> .../platform-mipsel/config/images/malta.config | 16 ++++------------
> .../config/images/vexpress-nor.config | 10 +++-------
> .../platform-v7a/config/images/vexpress.config | 15 +++------------
> .../config/images/vexpress-nor.config | 10 +++-------
> .../config/images/vexpress.config | 15 +++------------
> 7 files changed, 24 insertions(+), 74 deletions(-)
>
> diff --git a/configs/platform-mips/config/images/ar9331.config b/configs/platform-mips/config/images/ar9331.config
> index 07125bea4c88..8012cf0f839d 100644
> --- a/configs/platform-mips/config/images/ar9331.config
> +++ b/configs/platform-mips/config/images/ar9331.config
> @@ -1,7 +1,10 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 1G
> hdimage {
> align = 1M
> disk-signature = 0x74823abf
> + fill = true
> }
> partition root-A {
> image = root.ext2
> @@ -10,18 +13,7 @@ image @IMAGE@ {
> }
> partition data {
> partition-type = 0x83
> - size = 512M
> - }
> - /* genimage does not write zeros at the end of an image to save space,
> - * but the kernel in qemu will complain if the image file is smaller
> - * than the partition sizes. work around this by adding an empty
> - * partition at the end.
> - */
> - partition end {
> - image = /dev/null
> - size = 1M
> - offset = 1025M
> - in-partition-table = false
> + size = 511M
The last partition can have a 'fill = true' property if the image size is
explicitly specified. That's better than calculating it manually.
Michael
> }
> }
>
> diff --git a/configs/platform-mips/config/images/malta.config b/configs/platform-mips/config/images/malta.config
> index 07125bea4c88..8012cf0f839d 100644
> --- a/configs/platform-mips/config/images/malta.config
> +++ b/configs/platform-mips/config/images/malta.config
> @@ -1,7 +1,10 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 1G
> hdimage {
> align = 1M
> disk-signature = 0x74823abf
> + fill = true
> }
> partition root-A {
> image = root.ext2
> @@ -10,18 +13,7 @@ image @IMAGE@ {
> }
> partition data {
> partition-type = 0x83
> - size = 512M
> - }
> - /* genimage does not write zeros at the end of an image to save space,
> - * but the kernel in qemu will complain if the image file is smaller
> - * than the partition sizes. work around this by adding an empty
> - * partition at the end.
> - */
> - partition end {
> - image = /dev/null
> - size = 1M
> - offset = 1025M
> - in-partition-table = false
> + size = 511M
> }
> }
>
> diff --git a/configs/platform-mipsel/config/images/malta.config b/configs/platform-mipsel/config/images/malta.config
> index 07125bea4c88..8012cf0f839d 100644
> --- a/configs/platform-mipsel/config/images/malta.config
> +++ b/configs/platform-mipsel/config/images/malta.config
> @@ -1,7 +1,10 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 1G
> hdimage {
> align = 1M
> disk-signature = 0x74823abf
> + fill = true
> }
> partition root-A {
> image = root.ext2
> @@ -10,18 +13,7 @@ image @IMAGE@ {
> }
> partition data {
> partition-type = 0x83
> - size = 512M
> - }
> - /* genimage does not write zeros at the end of an image to save space,
> - * but the kernel in qemu will complain if the image file is smaller
> - * than the partition sizes. work around this by adding an empty
> - * partition at the end.
> - */
> - partition end {
> - image = /dev/null
> - size = 1M
> - offset = 1025M
> - in-partition-table = false
> + size = 511M
> }
> }
>
> diff --git a/configs/platform-v7a/config/images/vexpress-nor.config b/configs/platform-v7a/config/images/vexpress-nor.config
> index cc7edd28f628..58bbbfca404a 100644
> --- a/configs/platform-v7a/config/images/vexpress-nor.config
> +++ b/configs/platform-v7a/config/images/vexpress-nor.config
> @@ -1,6 +1,9 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 64M
> hdimage {
> partition-table-type = none
> + fill = true
> }
> partition barebox {
> image = barebox-vexpress-ca9.img
> @@ -19,13 +22,6 @@ image @IMAGE@ {
> in-partition-table = false
> size = 1M
> }
> - /* expand image size to 64 MiB; dummy size, results in 0 bytes length */
> - partition end {
> - image = /dev/null
> - offset = 64M
> - size = 1M
> - in-partition-table = false
> - }
> }
>
> /* vim: set tabstop=8 noexpandtab : */
> diff --git a/configs/platform-v7a/config/images/vexpress.config b/configs/platform-v7a/config/images/vexpress.config
> index 3aa64a4299ce..ba677fc3c946 100644
> --- a/configs/platform-v7a/config/images/vexpress.config
> +++ b/configs/platform-v7a/config/images/vexpress.config
> @@ -1,7 +1,10 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 256M
> hdimage {
> align = 1M
> disk-signature = 0x74823abf
> + fill = true
> }
> partition root-A {
> image = root.ext2
> @@ -17,18 +20,6 @@ image @IMAGE@ {
> partition-type = 0x83
> size = 63M
> }
> - /* genimage does not write zeros at the end of an image to save space,
> - * but the kernel in qemu will complain if the image file is smaller
> - * than the partition sizes, and qemu will complain if the SD card size
> - * is not a power of 2. work around this issues by adding an empty
> - * partition at the end.
> - */
> - partition end {
> - image = /dev/null
> - size = 1M
> - offset = 256M
> - in-partition-table = false
> - }
> }
>
> /* vim: set tabstop=8 noexpandtab : */
> diff --git a/configs/platform-v7a_noneon/config/images/vexpress-nor.config b/configs/platform-v7a_noneon/config/images/vexpress-nor.config
> index cc7edd28f628..58bbbfca404a 100644
> --- a/configs/platform-v7a_noneon/config/images/vexpress-nor.config
> +++ b/configs/platform-v7a_noneon/config/images/vexpress-nor.config
> @@ -1,6 +1,9 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 64M
> hdimage {
> partition-table-type = none
> + fill = true
> }
> partition barebox {
> image = barebox-vexpress-ca9.img
> @@ -19,13 +22,6 @@ image @IMAGE@ {
> in-partition-table = false
> size = 1M
> }
> - /* expand image size to 64 MiB; dummy size, results in 0 bytes length */
> - partition end {
> - image = /dev/null
> - offset = 64M
> - size = 1M
> - in-partition-table = false
> - }
> }
>
> /* vim: set tabstop=8 noexpandtab : */
> diff --git a/configs/platform-v7a_noneon/config/images/vexpress.config b/configs/platform-v7a_noneon/config/images/vexpress.config
> index d51a58be2893..5291cb31e385 100644
> --- a/configs/platform-v7a_noneon/config/images/vexpress.config
> +++ b/configs/platform-v7a_noneon/config/images/vexpress.config
> @@ -1,7 +1,10 @@
> image @IMAGE@ {
> + /* Note: qemu will complain if the SD card size is not a power of two */
> + size = 256M
> hdimage {
> align = 1M
> disk-signature = 0x74823abf
> + fill = true
> }
> partition root-A {
> image = root.ext2
> @@ -12,18 +15,6 @@ image @IMAGE@ {
> partition-type = 0x83
> size = 63M
> }
> - /* genimage does not write zeros at the end of an image to save space,
> - * but the kernel in qemu will complain if the image file is smaller
> - * than the partition sizes, and qemu will complain if the SD card size
> - * is not a power of 2. work around this issues by adding an empty
> - * partition at the end.
> - */
> - partition end {
> - image = /dev/null
> - size = 1M
> - offset = 256M
> - in-partition-table = false
> - }
> }
>
> /* vim: set tabstop=8 noexpandtab : */
> --
> 2.39.2
>
>
>
--
Pengutronix e.K. | |
Steuerwalder Str. 21 | http://www.pengutronix.de/ |
31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
next prev parent reply other threads:[~2023-10-11 6:51 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-10 19:16 [DistroKit] [PATCH 1/6] v7a: image-stm32mp: use 'partition-table-type' option instead of 'gpt' Roland Hieber
2023-10-10 19:16 ` [DistroKit] [PATCH 2/6] v7a: images: use 'partition-table-type' option instead of 'partition-table' Roland Hieber
2023-10-10 19:16 ` [DistroKit] [PATCH 3/6] v7a: barebox: pass root=mmcblkXpN instead of root=PARTUUID=… to kernel Roland Hieber
2023-10-10 19:16 ` [DistroKit] [PATCH 4/6] v7a: barebox: enable GUID Partition Table support Roland Hieber
2023-10-10 19:16 ` [DistroKit] [PATCH 5/6] images: remove qemu size hack Roland Hieber
2023-10-11 6:51 ` Michael Olbrich [this message]
2023-10-11 9:40 ` Roland Hieber
2023-10-11 9:45 ` Roland Hieber
2023-10-11 10:03 ` Roland Hieber
2023-10-10 19:16 ` [DistroKit] [PATCH 6/6] images: convert to GPT format Roland Hieber
2023-10-11 6:58 ` Michael Olbrich
2023-10-11 9:34 ` Roland Hieber
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=ZSZF43fZNs8TehiL@pengutronix.de \
--to=m.olbrich@pengutronix.de \
--cc=distrokit@pengutronix.de \
--cc=rhi@pengutronix.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