#1140461 spam wave of kernel: tpm tpm0: tpm_try_transmit: send(): error -62

#1140461#5
Date:
2026-06-21 01:38:06 UTC
From:
To:
I freshly installed Debian Stable on some PCs and those are getting
spammed by the kernel. Those are way frequent, pushing other logs in
journal.

Could you please tone this down, maybe stop writing to journal so
frequently?

This is not just my problem, because searching for it yields similar
reports;

-
https://forums.unraid.net/topic/192064-kernel-tpm-tpm0-tpm_try_transmit-send-error-62/
-
https://community.st.com/interface-and-connectivity-ics-52/linux-driver-for-tcg-tpm-i2c-drv-returning-error-62-after-chip-reset-through-gpio-131942
-
https://lists.debian.org/.internal/challenge.html?original=%2fdebian-kernel%2f2025%2f04%2fmsg00146.html
(BTW your "I challenge thee" loops forever until I press F5 on browser)

# lsmod | grep tpm_
tpm_infineon           16384  0

# cat /sys/class/tpm/tpm*/tpm_version_major 2>/dev/null
1

# journalctl --all|grep tpm|head -n 10
XX:00:06 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:00:18 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:00:30 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:00:42 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:00:54 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:01:06 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:01:18 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:01:30 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:01:42 kernel: tpm tpm0: tpm_try_transmit: send(): error -62
XX:01:54 kernel: tpm tpm0: tpm_try_transmit: send(): error -62

# dpkg --list|grep linux-
ii  linux-base                            4.12.1
      all          Linux image base package
ii  linux-image-6.12.74+deb13+1-amd64     6.12.74-2
      amd64        Linux 6.12 for 64-bit PCs (signed)
ii  linux-image-6.12.90+deb13.1-amd64     6.12.90-2
      amd64        Linux 6.12 for 64-bit PCs (signed)
ii  linux-image-amd64                     6.12.90-2
      amd64        Linux for 64-bit PCs (meta-package)
ii  linux-sysctl-defaults                 4.12.1
      all          default sysctl configuration for Linux

(I thought about blacklisting the 'tpm_infineon' using modprobe but I'm
not sure this is the good solution so I'm not doing it. I'm running
Debian using "Encrypted KVM" so if I disable tpm I might lose access)

#1140461#20
Date:
2026-06-21 06:58:47 UTC
From:
To:
Are you experiencing this on the VM host, or inside the VM? If inside
the VM and you're not using the TPM, consider removing the TPM from the
VM or try a Debian cloud kernel.

Greetings
Marc

#1140461#25
Date:
2026-06-23 00:21:51 UTC
From:
To:
I didn't write Virtual Machine at all - all of them are bare metal with
monitor (PC).

Here's another similar report:
-
https://forum.manjaro.org/t/repeated-tpm-try-transmit-send-error-62/172139/3
"using the infineon kernel module"

It seems this 'infineon' TPM have a problem.

If you can fix this on linux side, by not spamming journal, that'll be
appreciated by everyone who have similar symptom.

I _could_ disable TPM via BIOS, but I'm not sure I can do this because:
1. This PC's BIOS is old & it may not have TPM settings
2. I had an incident when I disabled TPM & could not boot anymore when I
was using Windows. Now I moved everything to Debian, I don't want to
touch TPM at all as it may break my Encrypted KVM...

#1140461#30
Date:
2026-06-23 00:34:57 UTC
From:
To:
For the temporary workaround - here's how I solved this.
I guess I need to run this script on cron's @reboot for now.
root:/tmp# lsmod | grep -i tpm
tpm_infineon           16384  0
root:/tmp# rmmod tpm_infineon
root:/tmp# journalctl --all --since="1 minute ago"
00:26:53 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:05 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:17 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:29 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:41 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
root:/tmp# journalctl --all --since="1 minute ago"
00:27:05 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:17 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:29 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:41 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:27:53 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62 <---
NEW!
root:/tmp# ls /sys/class/tpm/tpm0/device/driver
00:01  bind  uevent  unbind
root:/tmp# echo 00:01|tee /sys/class/tpm/tpm0/device/driver/unbind
00:01
root:/tmp# journalctl --all --since="1 minute ago"
00:28:53 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:29:05 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:29:17 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:29:30 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
00:29:42 sv kernel: tpm tpm0: tpm_try_transmit: send(): error -62
root:/tmp# journalctl --all --since="1 minute ago"
-- No entries --

#1140461#35
Date:
2026-06-23 08:53:36 UTC
From:
To:
[...]

I think there was some confusion because you are referring to "Encrypted
KVM" and KVM is the Linux kernel virtual machine facility.

Presumably that should be "Encrypted LVM"?

Ben.

#1140461#40
Date:
2026-06-24 16:30:41 UTC
From:
To:
Hello,

Alternatively you could pass

	module_blacklist=tpm_infineon

on the kernel commandline to get this worked around a bit cleaner.

Best regards
Uwe

#1140461#45
Date:
2026-06-24 16:34:46 UTC
From:
To:
Hello,

this is probably the same bug as https://bugs.debian.org/1139269 . For
that we (= Debian kernel team) wait for the reporter to provide a full
kernel log. So your full kernel log would help, too. (Without the
blacklist_module parameter that I suggested on the other end of this bug
thread.)

Best regards
Uwe

#1140461#50
Date:
2026-06-24 16:34:46 UTC
From:
To:
Hello,

this is probably the same bug as https://bugs.debian.org/1139269 . For
that we (= Debian kernel team) wait for the reporter to provide a full
kernel log. So your full kernel log would help, too. (Without the
blacklist_module parameter that I suggested on the other end of this bug
thread.)

Best regards
Uwe