Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that should interfere with TF-A) - Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs loaded via TFTP
Regards, Thomas Bonnefille
Hello Thomas,
Can you please share some logs of where you're seeing the hang and how you built tfa? Unfortunately, I have no way of reproducing what you're seeing since I have no access to a board. When the LTO patch was merged, all platforms in the CI were tested and confirmed working, so the problem is likely isolated to the platform's code. It's also likely that whatever went wrong has been a latent bug that LTO has uncovered as we've seen a few of these. I'll have a look when you post some logs but someone with access to the platform is more appropriate to debug this.
Thanks, Boyan ________________________________ From: Thomas Bonnefille via TF-A tf-a@lists.trustedfirmware.org Sent: 24 February 2026 6:02 PM To: tf-a@lists.trustedfirmware.org tf-a@lists.trustedfirmware.org Subject: [TF-A] Kernel freezes at boot (commit fa28b3a)
Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that should interfere with TF-A) - Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs loaded via TFTP
Regards, Thomas Bonnefille
Hi Boyan and Thomas,
Let me have a try on my side.
BR Jacky Bai From: Boyan Karatotev via TF-A tf-a@lists.trustedfirmware.org Sent: 2026年2月27日 16:59 To: tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello Thomas,
Can you please share some logs of where you're seeing the hang and how you built tfa? Unfortunately, I have no way of reproducing what you're seeing since I have no access to a board. When the LTO patch was merged, all platforms in the CI were tested and confirmed working, so the problem is likely isolated to the platform's code. It's also likely that whatever went wrong has been a latent bug that LTO has uncovered as we've seen a few of these. I'll have a look when you post some logs but someone with access to the platform is more appropriate to debug this.
Thanks, Boyan ________________________________ From: Thomas Bonnefille via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Sent: 24 February 2026 6:02 PM To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Subject: [TF-A] Kernel freezes at boot (commit fa28b3a)
Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that should interfere with TF-A) - Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs loaded via TFTP
Regards, Thomas Bonnefille
Hi Thomas,
I did not able to reproduce this issue on my side. Which GCC version are you using?
BR Jacky Bai From: Jacky Bai Sent: 2026年2月27日 17:08 To: Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: RE: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hi Boyan and Thomas,
Let me have a try on my side.
BR Jacky Bai From: Boyan Karatotev via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Sent: 2026年2月27日 16:59 To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org; Thomas Bonnefille <thomas.bonnefille@bootlin.commailto:thomas.bonnefille@bootlin.com> Subject: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello Thomas,
Can you please share some logs of where you're seeing the hang and how you built tfa? Unfortunately, I have no way of reproducing what you're seeing since I have no access to a board. When the LTO patch was merged, all platforms in the CI were tested and confirmed working, so the problem is likely isolated to the platform's code. It's also likely that whatever went wrong has been a latent bug that LTO has uncovered as we've seen a few of these. I'll have a look when you post some logs but someone with access to the platform is more appropriate to debug this.
Thanks, Boyan ________________________________ From: Thomas Bonnefille via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Sent: 24 February 2026 6:02 PM To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Subject: [TF-A] Kernel freezes at boot (commit fa28b3a)
Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that should interfere with TF-A) - Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs loaded via TFTP
Regards, Thomas Bonnefille
Hello everyone, thank you very much for your time.
Here are all the steps I did to produce this bug:
First my setup : I use the aarch64 cross-toolchain from https://toolchains.bootlin.com/ with gcc 14.3.0. I'm currently running it on Arch Linux with a kernel 6.18.13 (if it's important)
I got TF-A from github.com/ARM-software/arm-trusted-firmware on commit fa28b3afd. And built it with : `make PLAT=imx93 bl31` and copied the resulting `build/imx93/release/bl31.bin` in barebox/firmware/imx93-bl31.bin.
Regarding the version of barebox I'm using, it's an unreleased barebox version from my customer so I can't send it here without asking him first. But it is based on barebox 2025.09.0 with custom ddr timings, a PMIC driver, a SPI controller driver and a custom board file. From what I see, nothing here should mess with TF-A but if you suspect this unknown brick I can probably have a look to reduce it to the bare minimum and anonymize it to send it here.
In all cases I compiled it with the defconfig attached (I just removed the customer parts) with a simple `make -j24`
Regarding the kernel, I'm using a 6.18.13 mainline kernel built with Yocto and the attached defconfig and a basic rootfs built with Yocto too.
I attached the logs I'm seeing from power-up to the hang. After : ``` [ 1.426950] mmcblk0boot0: mmc0:0001 M08A11 4.00 MiB [ 1.434456] mmcblk0boot1: mmc0:0001 M08A11 4.00 MiB [ 1.441226] mmcblk0rpmb: mmc0:0001 M08A11 4.00 MiB, chardev (240:0) ``` there are no more logs and I can't interact with the console anymore, it just hangs indefinitely.
Using a TF-A version without this commit I'm able to get to a working shell.
Feel free to ask if you want more logs or if you want me to test some specific things.
Best regards, Thomas Bonnefille
On Wed Mar 4, 2026 at 3:06 AM CET, Jacky Bai wrote:
Hi Thomas,
I did not able to reproduce this issue on my side. Which GCC version are you using?
BR Jacky Bai From: Jacky Bai Sent: 2026年2月27日 17:08 To: Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: RE: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hi Boyan and Thomas,
Let me have a try on my side.
BR Jacky Bai From: Boyan Karatotev via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Sent: 2026年2月27日 16:59 To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org; Thomas Bonnefille <thomas.bonnefille@bootlin.commailto:thomas.bonnefille@bootlin.com> Subject: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello Thomas,
Can you please share some logs of where you're seeing the hang and how you built tfa? Unfortunately, I have no way of reproducing what you're seeing since I have no access to a board. When the LTO patch was merged, all platforms in the CI were tested and confirmed working, so the problem is likely isolated to the platform's code. It's also likely that whatever went wrong has been a latent bug that LTO has uncovered as we've seen a few of these. I'll have a look when you post some logs but someone with access to the platform is more appropriate to debug this.
Thanks, Boyan ________________________________ From: Thomas Bonnefille via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Sent: 24 February 2026 6:02 PM To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org> Subject: [TF-A] Kernel freezes at boot (commit fa28b3a)
Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that should interfere with TF-A)
- Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs loaded via TFTP
Regards, Thomas Bonnefille
Hi Thomas,
Please refer to this mail thread to have a try? From the test did on my side, the changes in the patch seems can also resolve the issue. https://review.trustedfirmware.org/c/TF-A/trusted-firmware-a/+/48633
BR Jacky Bai
-----Original Message----- From: Thomas Bonnefille thomas.bonnefille@bootlin.com Sent: 2026年3月6日 0:13 To: Jacky Bai ping.bai@nxp.com; Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: Re: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello everyone, thank you very much for your time.
Here are all the steps I did to produce this bug:
First my setup : I use the aarch64 cross-toolchain from https://toolchains.bootlin.com/ with gcc 14.3.0. I'm currently running it on Arch Linux with a kernel 6.18.13 (if it's important)
I got TF-A from github.com/ARM-software/arm-trusted-firmware on commit fa28b3afd. And built it with : `make PLAT=imx93 bl31` and copied the resulting `build/imx93/release/bl31.bin` in barebox/firmware/imx93-bl31.bin.
Regarding the version of barebox I'm using, it's an unreleased barebox version from my customer so I can't send it here without asking him first. But it is based on barebox 2025.09.0 with custom ddr timings, a PMIC driver, a SPI controller driver and a custom board file. From what I see, nothing here should mess with TF-A but if you suspect this unknown brick I can probably have a look to reduce it to the bare minimum and anonymize it to send it here.
In all cases I compiled it with the defconfig attached (I just removed the customer parts) with a simple `make -j24`
Regarding the kernel, I'm using a 6.18.13 mainline kernel built with Yocto and the attached defconfig and a basic rootfs built with Yocto too.
I attached the logs I'm seeing from power-up to the hang. After :
[ 1.426950] mmcblk0boot0: mmc0:0001 M08A11 4.00 MiB [ 1.434456] mmcblk0boot1: mmc0:0001 M08A11 4.00 MiB [ 1.441226] mmcblk0rpmb: mmc0:0001 M08A11 4.00 MiB, chardev (240:0)there are no more logs and I can't interact with the console anymore, it just hangs indefinitely.
Using a TF-A version without this commit I'm able to get to a working shell.
Feel free to ask if you want more logs or if you want me to test some specific things.
Best regards, Thomas Bonnefille
On Wed Mar 4, 2026 at 3:06 AM CET, Jacky Bai wrote:
Hi Thomas,
I did not able to reproduce this issue on my side. Which GCC version are you
using?
BR Jacky Bai From: Jacky Bai Sent: 2026年2月27日 17:08 To: Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: RE: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hi Boyan and Thomas,
Let me have a try on my side.
BR Jacky Bai From: Boyan Karatotev via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org
Sent: 2026年2月27日 16:59 To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org; Thomas Bonnefille
<thomas.bonnefille@bootlin.commailto:thomas.bonnefille@bootlin.com>
Subject: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello Thomas,
Can you please share some logs of where you're seeing the hang and how
you built tfa? Unfortunately, I have no way of reproducing what you're seeing since I have no access to a board. When the LTO patch was merged, all platforms in the CI were tested and confirmed working, so the problem is likely isolated to the platform's code. It's also likely that whatever went wrong has been a latent bug that LTO has uncovered as we've seen a few of these. I'll have a look when you post some logs but someone with access to the platform is more appropriate to debug this.
Thanks, Boyan ________________________________ From: Thomas Bonnefille via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org
Sent: 24 February 2026 6:02 PM To: tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.org
Subject: [TF-A] Kernel freezes at boot (commit fa28b3a)
Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that should
interfere with TF-A)
- Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs loaded via TFTP
Regards, Thomas Bonnefille
Hi Thomas, Boyan,
I think the root cause is clear for now. The fix in below commit is valid.
BR Jacky Bai
-----Original Message----- From: Jacky Bai Sent: 2026年3月9日 11:00 To: Thomas Bonnefille thomas.bonnefille@bootlin.com; Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org Subject: RE: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hi Thomas,
Please refer to this mail thread to have a try? From the test did on my side, the changes in the patch seems can also resolve the issue. https://review.trustedfirmware.org/c/TF-A/trusted-firmware-a/+/48633
BR Jacky Bai
-----Original Message----- From: Thomas Bonnefille thomas.bonnefille@bootlin.com Sent: 2026年3月6日 0:13 To: Jacky Bai ping.bai@nxp.com; Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: Re: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello everyone, thank you very much for your time.
Here are all the steps I did to produce this bug:
First my setup : I use the aarch64 cross-toolchain from https://toolchains.bootlin.com/ with gcc 14.3.0. I'm currently running it on Arch Linux with a kernel 6.18.13 (if it's important)
I got TF-A from github.com/ARM-software/arm-trusted-firmware on commit fa28b3afd. And built it with : `make PLAT=imx93 bl31` and copied the resulting `build/imx93/release/bl31.bin` in barebox/firmware/imx93-bl31.bin.
Regarding the version of barebox I'm using, it's an unreleased barebox version from my customer so I can't send it here without asking him first. But it is based on barebox 2025.09.0 with custom ddr timings, a PMIC driver, a SPI controller driver and a custom board file. From what I see, nothing here should mess with TF-A but if you suspect this unknown brick I can probably have a look to reduce it to the bare minimum and anonymize it to send it here.
In all cases I compiled it with the defconfig attached (I just removed the customer parts) with a simple `make -j24`
Regarding the kernel, I'm using a 6.18.13 mainline kernel built with Yocto and the attached defconfig and a basic rootfs built with Yocto too.
I attached the logs I'm seeing from power-up to the hang. After :
[ 1.426950] mmcblk0boot0: mmc0:0001 M08A11 4.00 MiB [ 1.434456] mmcblk0boot1: mmc0:0001 M08A11 4.00 MiB [ 1.441226] mmcblk0rpmb: mmc0:0001 M08A11 4.00 MiB, chardev (240:0)there are no more logs and I can't interact with the console anymore, it just hangs indefinitely.
Using a TF-A version without this commit I'm able to get to a working shell.
Feel free to ask if you want more logs or if you want me to test some specific things.
Best regards, Thomas Bonnefille
On Wed Mar 4, 2026 at 3:06 AM CET, Jacky Bai wrote:
Hi Thomas,
I did not able to reproduce this issue on my side. Which GCC version are you
using?
BR Jacky Bai From: Jacky Bai Sent: 2026年2月27日 17:08 To: Boyan Karatotev Boyan.Karatotev@arm.com; tf-a@lists.trustedfirmware.org; Thomas Bonnefille thomas.bonnefille@bootlin.com Subject: RE: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hi Boyan and Thomas,
Let me have a try on my side.
BR Jacky Bai From: Boyan Karatotev via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.or g
Sent: 2026年2月27日 16:59 To: tf-a@lists.trustedfirmware.org<mailto:tf-a@lists.trustedfirmware.org
;
Thomas Bonnefille
<thomas.bonnefille@bootlin.commailto:thomas.bonnefille@bootlin.com>
Subject: [TF-A] Re: Kernel freezes at boot (commit fa28b3a)
Hello Thomas,
Can you please share some logs of where you're seeing the hang and how
you built tfa? Unfortunately, I have no way of reproducing what you're seeing since I have no access to a board. When the LTO patch was merged, all platforms in the CI were tested and confirmed working, so the problem is likely isolated to the platform's code. It's also likely that whatever went wrong has been a latent bug that LTO has
uncovered as we've seen a few of these.
I'll have a look when you post some logs but someone with access to the platform is more appropriate to debug this.
Thanks, Boyan ________________________________ From: Thomas Bonnefille via TF-A <tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.or g
Sent: 24 February 2026 6:02 PM To: tf-a@lists.trustedfirmware.org<mailto:tf-a@lists.trustedfirmware.org
<tf-a@lists.trustedfirmware.orgmailto:tf-a@lists.trustedfirmware.or g
Subject: [TF-A] Kernel freezes at boot (commit fa28b3a)
Hello, I'm currently working on an NXP i.MX93 and I'm having issues with the commit fa28b3a adding, by default, link-time optimization.
When using a commit more recent than fa28b3a my kernel freezes at some point (after 1.2s to 1.8s from boot), there are no logs indicating a problem, just a freeze. It doesn't always appear at the same time in boot but I've never been able to reach a shell.
Using master with this commit reverted results in a working system.
If you want to reproduce the issue, I did produce it with :
- Mainline Barebox 2025.09 (slightly modified but nothing that
should interfere with TF-A)
- Mainline Linux 6.18.13 with a simple embedded Buildroot initramfs
loaded
via TFTP
Regards, Thomas Bonnefille
tf-a@lists.trustedfirmware.org