linux-qcom-next/linux-yocto-dev: enable more use cases #1180

koenkooi · 2025-10-31T12:50:42Z

Enable the following kernel features:

ZRAM, when configured by userspace, can provide a compressed swapspace backed by regular RAM. This generally increases the amount of available memory, avoiding the need for disk-backed swap. When it isn't configured by userspace, this does nothing.
In-kernel wifi infrastructure, this makes the iwd backend for connman/networkmanager/systemd work
Binfmt-misc, this makes podman, FEX and other software that wants transparent interpreter remapping work

lumag

Should it be also enabled for linux-yocto / linux-yocto-dev?
Also, please rebase to fix build issue.

koenkooi · 2025-11-01T09:01:11Z

Should it be also enabled for linux-yocto / linux-yocto-dev? Also, please rebase to fix build issue.

I've added it to linux-yocto-dev and rebased. Do we want this in linux-yocto non-dev as well? The current .scc is empty, so it seems to be out of sync already :/

lumag · 2025-11-01T09:28:05Z

Should it be also enabled for linux-yocto / linux-yocto-dev? Also, please rebase to fix build issue.

I've added it to linux-yocto-dev and rebased. Do we want this in linux-yocto non-dev as well? The current .scc is empty, so it seems to be out of sync already :/

The scc is there to pull out the rest of the files.

koenkooi · 2025-11-05T09:55:56Z

I've added more features, but left out linux-yocto-nondev till we sort out those configs in #1184 and #1186

lool · 2025-11-05T10:05:36Z

How likely are we to get these config changes accepted upstream?

koenkooi · 2025-11-05T10:16:01Z

How likely are we to get these config changes accepted upstream?

I am unsure, Linus himself has been pushing back against various additions to the defconfigs saying distros should enable and test the features they want. I don't know where he draws the line and how much attention he pays to this.

koenkooi · 2025-11-05T10:20:29Z

How likely are we to get these config changes accepted upstream?

This sums up the situation quite nicely, https://lwn.net/Articles/1026337/, especially this excerpt:

Torvalds, in turn, strongly told Molnar to stop this work, saying that the default configuration should be for "normal people". Options that are useful to cloud providers (such as the virtualization subsystems) should not be enabled, he said. The fact that all distributors enable a specific option is also not, in his mind, an argument for enabling that option in the default configuration.

lool · 2025-11-05T11:47:44Z

How likely are we to get these config changes accepted upstream?

This sums up the situation quite nicely, https://lwn.net/Articles/1026337/, especially this excerpt:

Torvalds, in turn, strongly told Molnar to stop this work, saying that the default configuration should be for "normal people". Options that are useful to cloud providers (such as the virtualization subsystems) should not be enabled, he said. The fact that all distributors enable a specific option is also not, in his mind, an argument for enabling that option in the default configuration.

I had Yocto as an upstream in mind here :)

lumag · 2025-11-07T17:57:47Z

recipes-kernel/linux/linux-qcom-next/configs/qcom.cfg

 CONFIG_ZRAM_MULTI_COMP=y
+
+# Wifi
+CONFIG_ASYMMETRIC_KEY_TYPE=y


I think most of these should be built as modules already. Could you please check / confirm?

I'll check later today, this section of the config is a copy/paste of what iwd told me to enable. I can turn all tristate options in =m, but I'd like to keep this list of options to make it easier to lift it into different kernels, that won't have anything enabled yet.

I'd rather not keep the full list here. It is ridiculously long and verbose. Most of the options are enabled by default. If you list all of them, for me it means that all of them are set incorrectly in the qcom-next config or in the defconfig.

The options that are enabled in linux-qcom-next in its current form are these:

CONFIG_ASYMMETRIC_KEY_TYPE=y CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE=y CONFIG_CFG80211_CRDA_SUPPORT=y CONFIG_CRYPTO_AES=y CONFIG_CRYPTO_CBC=m CONFIG_CRYPTO_CMAC=m CONFIG_CRYPTO_DES=m CONFIG_CRYPTO_ECB=y CONFIG_CRYPTO_HMAC=y CONFIG_CRYPTO_MD5=m CONFIG_CRYPTO_RSA=y CONFIG_CRYPTO_SHA1=m CONFIG_CRYPTO_SHA256=m CONFIG_CRYPTO_SHA512=y CONFIG_PKCS7_MESSAGE_PARSER=y CONFIG_X509_CERTIFICATE_PARSER=y

Which means that this specific kernel only needs the following:

CONFIG_CRYPTO_ARC4=m CONFIG_CRYPTO_MD4=m CONFIG_CRYPTO_USER_API_AEAD=m CONFIG_CRYPTO_USER_API_HASH=m CONFIG_CRYPTO_USER_API_SKCIPHER=m CONFIG_KEY_DH_OPERATIONS=y CONFIG_PKCS8_PRIVATE_KEY_PARSER=y

I'd still like to keep the complete list, since that makes it easier and less error-prone to copy this between kernels (linux-qcom-next, linux-yocto-dev, linux-yocto and vendor kernels).

I also don't trust upstream linux to not regress defconfigs, especially for things we currently don't test.

If you want a comprehensive list, please also enable ARM-specific crypto acceleration options.
At the same time, enabling obsolete & broken algorithms (DES, MD4, MD5) sounds very strange to me.

EAP-PEAP with MsCHAPv2 still uses MD4, EAP-MD5 uses, well MD5 and 3DES is sadly still allowed as an alternative to AES when using 802.1x :(

I'll have a look at the ARM-specific crypto options.

lumag · 2025-11-10T10:35:42Z

recipes-kernel/linux/linux-qcom-next/configs/qcom.cfg

 CONFIG_VETH=m
+
+# ZRAM
+CONFIG_ZRAM=y


Likewise. Is CONFIG_ZRAM=y not enough?

ZRAM, when configured by userspace, can provide a compressed swapspace backed by regular RAM. This generally increases the amount of available memory, avoiding the need for disk-backed swap. When it isn't configured by userspace, this does nothing. Signed-off-by: Koen Kooi <[email protected]>

This also enables accelerated crypto drivers for Arm, the wifi features use a subset of these. When starting iwd: Starting Wireless service... No HMAC(SHA1) support found No HMAC(MD5) support found No CMAC(AES) support found No HMAC(SHA256) support not found No HMAC(SHA512) support found, certain TLS connections might fail DES support not found AES support not found No CBC(DES3_EDE) support found, certain TLS connections might fail No CBC(AES) support found, WPS will not be available No Diffie-Hellman support found, WPS will not be available The following options are missing in the kernel: CONFIG_CRYPTO_USER_API_HASH CONFIG_CRYPTO_USER_API_SKCIPHER CONFIG_KEY_DH_OPERATIONS CONFIG_CRYPTO_ECB CONFIG_CRYPTO_MD5 CONFIG_CRYPTO_CBC CONFIG_CRYPTO_SHA256 CONFIG_CRYPTO_AES CONFIG_CRYPTO_DES CONFIG_CRYPTO_CMAC CONFIG_CRYPTO_HMAC CONFIG_CRYPTO_SHA512 CONFIG_CRYPTO_SHA1 The following optimized implementations might be available: CONFIG_CRYPTO_SHA1_SSSE3 CONFIG_CRYPTO_AES_NI_INTEL CONFIG_CRYPTO_SHA512_SSSE3 CONFIG_CRYPTO_AES_X86_64 CONFIG_CRYPTO_DES3_EDE_X86_64 CONFIG_CRYPTO_SHA256_SSSE3 iwd.service: Main process exited, code=exited, status=1/FAILURE Signed-off-by: Koen Kooi <[email protected]> SQUASH ME: linux-yocto-dev: wifi, part 2 Signed-off-by: Koen Kooi <[email protected]>

This also enables accelerated crypto drivers for Arm, the wifi features use a subset of these. When starting iwd: Starting Wireless service... No HMAC(SHA1) support found No HMAC(MD5) support found No CMAC(AES) support found No HMAC(SHA256) support not found No HMAC(SHA512) support found, certain TLS connections might fail DES support not found AES support not found No CBC(DES3_EDE) support found, certain TLS connections might fail No CBC(AES) support found, WPS will not be available No Diffie-Hellman support found, WPS will not be available The following options are missing in the kernel: CONFIG_CRYPTO_USER_API_HASH CONFIG_CRYPTO_USER_API_SKCIPHER CONFIG_KEY_DH_OPERATIONS CONFIG_CRYPTO_ECB CONFIG_CRYPTO_MD5 CONFIG_CRYPTO_CBC CONFIG_CRYPTO_SHA256 CONFIG_CRYPTO_AES CONFIG_CRYPTO_DES CONFIG_CRYPTO_CMAC CONFIG_CRYPTO_HMAC CONFIG_CRYPTO_SHA512 CONFIG_CRYPTO_SHA1 The following optimized implementations might be available: CONFIG_CRYPTO_SHA1_SSSE3 CONFIG_CRYPTO_AES_NI_INTEL CONFIG_CRYPTO_SHA512_SSSE3 CONFIG_CRYPTO_AES_X86_64 CONFIG_CRYPTO_DES3_EDE_X86_64 CONFIG_CRYPTO_SHA256_SSSE3 iwd.service: Main process exited, code=exited, status=1/FAILURE Signed-off-by: Koen Kooi <[email protected]>

Signed-off-by: Koen Kooi <[email protected]>

Only half were enabled, enable them all. Signed-off-by: Koen Kooi <[email protected]>

quaresmajose · 2025-11-24T11:13:56Z

I've added more features, but left out linux-yocto-nondev till we sort out those configs in #1184 and #1186

The #1184 was closed and so I think you should also include the linux-yocto-nondev configs.

github-actions · 2025-11-24T13:15:59Z

Test run workflow

Test jobs for commit `5663248`

Job 119063 on dragonboard-410c - boot test (dragonboard-410c) 19635137398
Job 119057 on qcs6490 - boot test (qcs6490-rb3gen2-core-kit) 19635137398
Job 119062 on qcs6490 - boot test (qcs6490-rb3gen2-core-kit) 19635137398
Job 119058 on qcs9100-rb8 - boot test (iq-9075-evk) 19635137398
Job 119067 on qrb2210-rb1 - boot test (qrb2210-rb1-core-kit) 19635137398
Job 119065 on dragonboard-820c - boot test (dragonboard-820c) 19635137398
Job 119064 on qcs8300 - boot test (qcs8300-ride-sx) 19635137398
Job 119061 on qcs9100 - boot test (qcs9100-ride-sx) 19635137398
Job 119060 on qrb2210-rb1 - boot test (qrb2210-rb1-core-kit) 19635137398
Job 119068 on dragonboard-410c - boot test (dragonboard-410c) 19635137398
Job 119059 on qcs9100-rb8 - boot test (iq-9075-evk) 19635137398
Job 119069 on qcs8300 - boot test (qcs8300-ride-sx) 19635137398
Job 119066 on dragonboard-820c - boot test (dragonboard-820c) 19635137398
Job 119070 on qcs9100 - boot test (qcs9100-ride-sx) 19635137398

test-reporting-app · 2025-11-24T13:16:13Z

Test Results

14 files ±0 27 suites - 1 45m 31s ⏱️ + 11m 31s
52 tests +2 48 ✅ - 2 0 💤 ±0 4 ❌ +4
300 runs - 4 296 ✅ - 8 0 💤 ±0 4 ❌ +4

For more details on these failures, see this check.

Results for commit 5663248. ± Comparison against base commit edaaaf0.

koenkooi requested review from lumag, ndechesne, ricardosalveti, sbanerjee-quic and vkraleti as code owners October 31, 2025 12:50

lumag reviewed Oct 31, 2025

View reviewed changes

koenkooi force-pushed the koen/qcom-next-zram branch from 396edd4 to 3a9b16a Compare November 1, 2025 08:58

koenkooi force-pushed the koen/qcom-next-zram branch from 3a9b16a to 37dfcd3 Compare November 5, 2025 09:52

koenkooi changed the title ~~linux-qcom-next: enable ZRAM~~ linux-qcom-next/linux-yocto-dev: enable more use cases Nov 5, 2025

koenkooi requested a review from lumag November 5, 2025 09:56

lumag previously approved these changes Nov 6, 2025

View reviewed changes

lumag reviewed Nov 7, 2025

View reviewed changes

lumag reviewed Nov 10, 2025

View reviewed changes

recipes-kernel/linux/linux-qcom-next/configs/qcom.cfg

CONFIG_VETH=m

# ZRAM

CONFIG_ZRAM=y

Copy link

Contributor

lumag Nov 10, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Likewise. Is CONFIG_ZRAM=y not enough?

koenkooi added 6 commits November 24, 2025 11:38

linux-qcom-next: enable BINFMT_MISC to prevent podman crash

e87be67

Signed-off-by: Koen Kooi <[email protected]>

linux-qcom-next: enable all LED triggers

5663248

Only half were enabled, enable them all. Signed-off-by: Koen Kooi <[email protected]>

koenkooi dismissed lumag’s stale review via 5663248 November 24, 2025 10:51

koenkooi force-pushed the koen/qcom-next-zram branch from 37dfcd3 to 5663248 Compare November 24, 2025 10:51

linux-qcom-next/linux-yocto-dev: enable more use cases #1180

Are you sure you want to change the base?

linux-qcom-next/linux-yocto-dev: enable more use cases #1180

Uh oh!

Conversation

koenkooi commented Oct 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lumag left a comment

Choose a reason for hiding this comment

Uh oh!

koenkooi commented Nov 1, 2025

Uh oh!

lumag commented Nov 1, 2025

Uh oh!

koenkooi commented Nov 5, 2025

Uh oh!

lool commented Nov 5, 2025

Uh oh!

koenkooi commented Nov 5, 2025

Uh oh!

koenkooi commented Nov 5, 2025

Uh oh!

lool commented Nov 5, 2025

Uh oh!

lumag Nov 7, 2025

Choose a reason for hiding this comment

Uh oh!

koenkooi Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

lumag Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

koenkooi Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

lumag Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

koenkooi Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

lumag Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

quaresmajose commented Nov 24, 2025

Uh oh!

github-actions bot commented Nov 24, 2025

Test run workflow

Test jobs for commit 5663248

Uh oh!

test-reporting-app bot commented Nov 24, 2025

Test Results

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

koenkooi commented Oct 31, 2025 •

edited

Loading

Test jobs for commit `5663248`