Scarthgap #160
Open
Scarthgap #160
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Ferry Toth <[email protected]>
This flag is not standard for Yocto and for "Langdale" breaks building libgcc-initial: unwind.inc:141:1: sorry, unimplemented: Stack realignment not supported with '__builtin_eh_return' As it only causes unnecessary overhead drop it globally. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Support for building Yocto under windows is not available anyway Signed-off-by: Ferry Toth <[email protected]>
This fork of ninja it participate in the make jobserver pool. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
While here rename wifi firmware to 'brcmfmac43340-sdio.Intel Corporation-Merrifield.bin' as kernel prefers. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
It appears HOSTGCC is ignored which should be calling gcc-native. So -fcanon-prefix-map is not always supported as it is introduced in gcc 13. Override LDFLAGS by BUILD_LDFLAGS to suppress it. Signed-off-by: Ferry Toth <[email protected]>
Sometimes the non-blocking writer is closed. Open one to make there is at least one available. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
This applies only to 5.69 as used in nanbield. This fixes our bluetooth_bd_addr.sh that loads the correct public-addr into the BCM43341B0 bluetooth module. Without a valid public-addr the kernel driver will not initialize. Signed-off-by: Ferry Toth <[email protected]>
This make bash goto begin/end line with Home/End and move word forward/backeward with c-arrowleft/c-arrowright. Works on both Gtkterm and xterm. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
The script looks for files in build_dir, print it so the user is aware of the location. Signed-off-by: Ferry Toth <[email protected]>
U-boot has picked up detection of uncommited changes from Linux. Switch to using git instead of quilt to apply patches, and disable localversion to prevent the version to be extended with the last patches hash. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
xlt2tpd is pulled in by connman-vpn but as it is not correctly configured it errors and the service won't start. This append makes it a systemd service and then disables that. If needed in the future configuration bits should be added here. Signed-off-by: Ferry Toth <[email protected]>
The shell script broke long ago due to the change to btrfs. This fix should restore the original behavior by checking the free space on /. This is assuming /var/log is on / which is currently the case. Signed-off-by: Ferry Toth <[email protected]>
In our default configuration the journal is not persistant so this issue remains under the radar. However when making the journal persistant (by `mkdir /var/log/journal`) it fill up on each reboot. In the original behavior the journal got shutdown before reboot. That got dropped by 6724d49 when fixing reboot params behavior. Signed-off-by: Ferry Toth <[email protected]>
It appears somewhere since Dunfell selection of init manager was simplified, but we missed that. As result SysV initscripts are also being installed. Now it seems these scripts are conflicting systemd in some cases (umountfs and umountnfs). Setting INIT_MANAGER to systemd fixes this and also sets certain DISTRO_FEATURES that we do no longer need to explicitely set. Signed-off-by: Ferry Toth <[email protected]>
Currently makeing a snapshot of a dir with a subvolume in it leave a empty directory with the name of the subvolume. Delete it. While here, current filesystem hierarchy has an additional /srv dir, that should be removed from the top level sub volume. Signed-off-by: Ferry Toth <[email protected]>
Yocto switched to using Ninja by default, however that is broken for xfstk Signed-off-by: Ferry Toth <[email protected]>
…ount udevd stubbornly holds files open on /usr/lib/modules causing umounting to fail cleanly. This kills the sockets that trigger udevd to (re)start then stops systemd-udevd. Not the nicest solution. Suggestions for improvement welcome. Signed-off-by: Ferry Toth <[email protected]>
errors - show lines marked as error in the journal warnings - show lines marked as warning in the journal Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
This silences a warning: systemd-udevd.service: ProtectHostname=yes is configured, but the kernel does not support UTS namespaces, ignoring namespace setup. This also bring support for other name spaces (TIME, IPC, PID, NET) It supposedly brings security advantages hopefully without substantial costs. Signed-off-by: Ferry Toth <[email protected]>
As PREEMPT-RT is now in mainline we can enable it with a DISTRO_FEATURES
("rt"). The extra recipe was to distinguish between special SRC_URI which
is no longer needed (for X86).
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Move altboot.sh to the proper location. While here expand the syntax to make it suitable for variant (preempt-rt) kernels with the same version number. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Patch refresh for scarthgap Signed-off-by: Ferry Toth <[email protected]>
Need to wait a bit longer Signed-off-by: Ferry Toth <[email protected]>
After flashall likely the usb cable is still connected so if the host has network sharing configured we might be able to create a cdc (network over usb) connection. Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
Signed-off-by: Ferry Toth <[email protected]>
andy-shev
reviewed
Feb 6, 2025
| @@ -21,6 +21,7 @@ | |||
| NewArgv.append(i) | |||
|
|
|||
| if seen: | |||
| w = os.open(fifoname, os.O_WRONLY|os.O_NONBLOCK) | |||
There was a problem hiding this comment.
But is this w used anywhere? You can run pylint to see the warning it issues on this line (of course it may be more on the rest).
andy-shev
requested changes
Feb 6, 2025
|
|
||
| # Disable sysvinit for recipes with systemd support | ||
| DISTRO_FEATURES_BACKFILL_CONSIDERED += "sysvinit" | ||
| VIRTUAL-RUNTIME_initscripts = "" | ||
| VIRTUAL-RUNTIME_init_manager = "systemd" | ||
|
|
||
| # This builds tools and python3 bindings for libgpiod | ||
| PREFERRED_VERSION_libgpiod = "2.0" |
There was a problem hiding this comment.
Shouldn't this line be in one of the previous patches (WRT libgpiod removal)?
setup.sh
Outdated
| @@ -320,7 +320,8 @@ COPYLEFT_LICENSE_INCLUDE = 'GPL* LGPL*' | |||
| echo "Cloning meta-qt5 layer to ${oe_dir} directory from local cache" | |||
| git clone ${my_dl_dir}/meta-qt5-mirror.git meta-qt5 | |||
| cd ${oe_dir} | |||
| git checkout ${yocto_tag} | |||
| # git checkout ${yocto_tag} | |||
| @@ -7,7 +7,7 @@ | |||
|
|
|||
| dirs="bin boot dev etc home lib media mnt opt proc run sbin sketch sys tmp usr var" | |||
| deploy_dir=tmp/deploy/images/edison | |||
| image_name=edison-image-edison | |||
| image_name=edison-image-edison.rootfs | |||
There was a problem hiding this comment.
Commit message doesn't explain why. What will this help with?
| @@ -70,7 +70,7 @@ SRC_URI:append = " file://0044-REVERTME-usb-dwc3-gadget-skip-endpoints-ep-18-in- | |||
| SRC_URI:append = " file://0001-8250_mid-arm-rx-dma-on-all-ports-with-dma-continousl.patch" | |||
| SRC_URI:append = " file://0001a-serial-8250_dma-use-linear-buffer-for-transmit.patch" | |||
| SRC_URI:append = " file://0001-serial-8250_port-when-using-DMA-do-not-split-writes-.patch" | |||
| SRC_URI:append = " file://0001-usb-dwc3-core-Fix-dwc3_core_soft_reset-before-anythi.patch" | |||
| SRC_URI:append = " file://0001a-usb-dwc3-core-Fix-dwc3_core_soft_reset-before-anythi.patch" | |||
There was a problem hiding this comment.
Yeah, hopefully we will have eventually proper solution in v6.15-rc1.
| StandardOutput=journal+console | ||
| StandardError=journal | ||
| # Redirect stdin to ensure btmgmt doesn't hang | ||
| StandardInput=tty |
There was a problem hiding this comment.
Yeah, it would be nice to understand the root cause. I believe systemd has another (out of zillion) configuration parameter for such a case (like you don't need to specify tty explicitly). But it's all based on my pure speculation.
| @@ -19,6 +19,7 @@ TARGET_CC_ARCH += "${LDFLAGS}" | |||
| EXTRA_OECMAKE += " -DWARNING_AS_ERROR=OFF -DOE_QMAKE_PATH_EXTERNAL_HOST_BINS='${OE_QMAKE_PATH_EXTERNAL_HOST_BINS}'" | |||
|
|
|||
| OECMAKE_TARGET_COMPILE = "xfstk-dldr-solo" | |||
| OECMAKE_GENERATOR = "Unix Makefiles" | |||
There was a problem hiding this comment.
AFAIU this doesn't need to be handled in xFSTK project itself, correct?
|
|
||
| Current u-boot recipe in kirkstone fails during do_configure because | ||
| Current u-boot recipe in scarthgap fails during do_configure because |
There was a problem hiding this comment.
Hmm... Shouldn't be this fixed in the Yocto Poky? Or did I get this incorrectly?
| @@ -74,5 +74,5 @@ echo 120 > configs/c.1/MaxPower | |||
|
|
|||
| # Activate gadgets | |||
| echo dwc3.0.auto > UDC | |||
| sleep 1 | |||
There was a problem hiding this comment.
Basically you need to watch for a file in sysfs to appear (but which file, I don't know, perhaps the one that gives a new interface under net?).
# Wait until the interface has come up
for pass in $(seq 3); do
for eth in $(find /sys/class/net -type l -name 'eth*' -exec basename {} \;); do
| win32) | ||
| extra_conf="SDKMACHINE = \"i686-mingw32\"" | ||
| ;; | ||
| win64) |
There was a problem hiding this comment.
Does anybody try to build it under WSL2? I am just curious...
| install -v -d ${D}/lib/firmware/brcm/ | ||
| install -m 0755 edison-firmware/broadcom_cws/wlan/firmware/bcmdhd_aob.cal_4334x_b0 ${D}/lib/firmware/brcm/brcmfmac43340-sdio-fr.txt | ||
| install -m 0755 edison-firmware/broadcom_cws/wlan/firmware/bcmdhd.cal_4334x_b0 '${D}/lib/firmware/brcm/brcmfmac43340-sdio.Intel Corporation-Merrifield.txt' | ||
| install -m 0755 linux-firmware/cypress/cyfmac43340-sdio.bin ${D}/lib/firmware/brcm/brcmfmac43340-sdio.bin |
There was a problem hiding this comment.
The Cypress FW are available in linux-firmware repository, I'm wondering why we don't use them directly.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This updates from Yocto Kirkstone -> Langdale -> Mickledore -> Nanbield -> Scarthgap which is a LTS .
What's new is here.