| View previous topic :: View next topic |
| Author |
Message |
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
 Posted: Mon Aug 28, 2023 4:01 pm Post subject: [SOLVED]SD Card not discovered on boot, ok if re-inserted. |
 |
|
UPD:
Solved, see this post.
Hardware is Asus Chromebook CX1400CNA. (C423 Rabbid with MrChromebox-4.20.0 (05/15/2023) firmware onboard).
------------------------------------------------------------------------
Hi all! I recently bumped into the following problem - my sd card doen't get discovered if it is inserted into my laptop prior to boot, but everything is ok if I re-insert it after the system is up and running.
I've got eMMC memory, and it looks pretty ok:
| Code: | Aug 28 12:13:18 xk-crbook kernel: mmc1: SDHCI controller on PCI [0000:00:1b.0] using ADMA 64-bit
Aug 28 12:13:17 xk-crbook systemd-fsck[1223]: /dev/mmcblk0p1: 27 files, 23377/130812 clusters
Aug 28 12:13:17 xk-crbook kernel: Adding 2097148k swap on /dev/mmcblk0p2. Priority:-2 extents:1 across:2097148k SSFS
Aug 28 12:13:16 xk-crbook systemd-fsck[1161]: /dev/mmcblk0p3: clean, 612994/3661824 files, 6593164/14624256 blocks
Aug 28 12:13:16 xk-crbook kernel: EXT4-fs (mmcblk0p3): re-mounted. Quota mode: none.
Aug 28 12:13:16 xk-crbook kernel: EXT4-fs (mmcblk0p3): mounted filesystem with ordered data mode. Quota mode: none.
Aug 28 12:13:16 xk-crbook kernel: mmcblk0rpmb: mmc0:0001 TX2964 4.00 MiB, chardev (243:0)
Aug 28 12:13:16 xk-crbook kernel: mmcblk0boot1: mmc0:0001 TX2964 4.00 MiB
Aug 28 12:13:16 xk-crbook kernel: mmcblk0boot0: mmc0:0001 TX2964 4.00 MiB
Aug 28 12:13:16 xk-crbook kernel: mmcblk0: p1 p2 p3
Aug 28 12:13:16 xk-crbook kernel: mmcblk0: mmc0:0001 TX2964 58.3 GiB
Aug 28 12:13:16 xk-crbook kernel: mmc0: new HS400 MMC card at address 0001
Aug 28 12:13:16 xk-crbook kernel: mmc0: SDHCI controller on PCI [0000:00:1c.0] using ADMA 64-bit |
But when it comes to the sd card the only message I get is the following:
| Code: | | Aug 28 12:13:18 xk-crbook kernel: mmc1: SDHCI controller on PCI [0000:00:1b.0] using ADMA 64-bit |
Later, when I re-insert the card similar messages appear:
| Code: | Aug 28 12:16:24 xk-crbook kernel: mmcblk1: p1
Aug 28 12:16:24 xk-crbook kernel: mmcblk1: mmc1:59b4 SDU1 58.6 GiB
Aug 28 12:16:24 xk-crbook kernel: mmc1: new ultra high speed SDR104 SDXC card at address 59b4 |
I would like the card to work from the very boot and be used as a second storage.
Could you please suggest possible steps to solve this, or at least a general direction?
Kernel config has the following:
| Code: | cat /usr/src/linux/.config | grep MMC
CONFIG_PCI_MMCONFIG=y
CONFIG_MMCONF_FAM10H=y
CONFIG_MMC=y
CONFIG_MMC_BLOCK=y
CONFIG_MMC_BLOCK_MINORS=8
# CONFIG_MMC_TEST is not set
# CONFIG_MMC_CRYPTO is not set
# MMC/SD/SDIO Host Controller Drivers
# CONFIG_MMC_DEBUG is not set
CONFIG_MMC_SDHCI=y
CONFIG_MMC_SDHCI_IO_ACCESSORS=y
CONFIG_MMC_SDHCI_PCI=y
# CONFIG_MMC_RICOH_MMC is not set
CONFIG_MMC_SDHCI_ACPI=y
CONFIG_MMC_SDHCI_PLTFM=y
# CONFIG_MMC_SDHCI_F_SDH30 is not set
# CONFIG_MMC_WBSD is not set
# CONFIG_MMC_TIFM_SD is not set
# CONFIG_MMC_SPI is not set
# CONFIG_MMC_CB710 is not set
# CONFIG_MMC_VIA_SDMMC is not set
# CONFIG_MMC_VUB300 is not set
# CONFIG_MMC_USHC is not set
# CONFIG_MMC_USDHI6ROL0 is not set
CONFIG_MMC_CQHCI=y
CONFIG_MMC_HSQ=m
# CONFIG_MMC_TOSHIBA_PCI is not set
# CONFIG_MMC_MTK is not set
# CONFIG_MMC_SDHCI_XENON is not set
CONFIG_PHY_INTEL_LGM_EMMC=y
|
Last edited by dk0393 on Tue Sep 05, 2023 11:56 am; edited 2 times in total |
|
| Back to top |
|
 |
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Mon Aug 28, 2023 5:05 pm Post subject: Re: SD Card is not discovered on boot, ok if re-inserted. |
|
|
dk0393,
Welcome to Gentoo Forums ! 
| dk0393 wrote: | I would like the card to work from the very boot and be used as a second storage.
Could you please suggest possible steps to solve this, or at least a general direction? |
Maybe you dont have all necessary modules configured static <*> in your kernel. You might check this with a "lsmod". If you see here any modules your MMC needs (also filesystem; ext4) you should change them from <M> to <*> in your kernel configuration.
(Reason is: Kernel initialize all <M>odules after access to its root partition; only built-in modules will be initialized before kernel has root access; If you want your root partition ON this sd card you surely will need also "rootwait" as kernel command line parameter). |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Mon Aug 28, 2023 6:08 pm Post subject: |
|
|
Hi, pietinger!
Thanks for the fast reply.
I tried to make kernel as monolythic as possible, lsmod output is the same both before and after re-inserting sd card (so, I didn't forget to make anything built-in), thus I may presume the problem might be somewhere deeper. The lspci output is:
| Code: |
...
00:1b.0 SD Host controller: Intel Corporation Celeron N3350/Pentium N4200/Atom E3900 Series SDXC/MMC Host Controller (rev 0b)
...
|
Just to clarify... The hardware is Asus chromebook cx1400, I heard they have multiple problems here and there, maybe this could be related to the problem. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Mon Aug 28, 2023 6:23 pm Post subject: |
|
|
| dk0393 wrote: | | [...] lsmod output is the same both before and after re-inserting sd card [...] |
That says nothing. Please give us the output of your "lsmod".
| dk0393 wrote: | | [...] I may presume the problem might be somewhere deeper. |
Maybe ... then we would need:
- Output of "lspci -k"
- Output of "dmesg" (after systemstart)
- Your current kernel .config
(please use wgetpaste for big files => https://wiki.gentoo.org/wiki/Wgetpaste) |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Tue Aug 29, 2023 8:20 am Post subject: |
|
|
| pietinger wrote: |
Maybe ... then we would need:
- Output of "lspci -k"
- Output of "dmesg" (after systemstart)
- Your current kernel .config
(please use wgetpaste for big files => https://wiki.gentoo.org/wiki/Wgetpaste) |
Here:
kernel config
dmesg
lspci -k
As a temporary solution I'm trying to figure out how to power off and on again the cardreader module before switch_root.
UPD:
Didn't find a way to power on/off, but the remove/rescan helps:
| Code: | echo "1" > /sys/bus/pci/devices/0000\:00\:1b.0/remove
echo "1" > /sys/bus/pci/rescan |
|
|
| Back to top |
|
|
szatox
Advocate
Joined: 27 Aug 2013
Posts: 3481
|
| Posted: Tue Aug 29, 2023 10:59 am Post subject: |
|
|
| Quote: | | SD Card is not discovered on boot, ok if re-inserted. |
"Works after hotplug but not after reboot" looks like udev to me.
You can trigger udev events manually with udevadm trigger -c add|change. Check out the manual for more options (and for ways to limit the scope of trigger if you need it)
Also, I wonder if there is any easy way to get a debug log from udev after running a trigger, it could help with fixing the problem with X not working after reboot. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Tue Aug 29, 2023 11:25 am Post subject: |
|
|
dk0393,
I have checked your kernel configuration ... and missed some lines .. these lines are not important, but it could indicate that you have edited the .config once. If this is true, please read:
https://wiki.gentoo.org/wiki/User:Pietinger/Tutorials/Manual_kernel_configuration#Never_edit_.config
I miss also your output of "lsmod" ... 
But I found a little bit:
| Code: | 1)
CONFIG_EXPERT=y
2)
CONFIG_MTRR_SANITIZER_ENABLE_DEFAULT=0
3)
CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
# CONFIG_CPU_FREQ_DEFAULT_GOV_SCHEDUTIL is not set
CONFIG_CPU_FREQ_GOV_PERFORMANCE=y
CONFIG_CPU_FREQ_GOV_POWERSAVE=y
4)
# CONFIG_SYSFB is not set
5)
# CONFIG_SCSI_PROC_FS is not set
6)
CONFIG_BLK_DEV_SD=m
CONFIG_BLK_DEV_SR=m
# CONFIG_SCSI_CONSTANTS is not set
7)
# CONFIG_SCSI_SPI_ATTRS is not set
8)
# CONFIG_ATA is not set
9)
CONFIG_I2C_SMBUS=m
CONFIG_I2C_ALGOBIT=m
10)
CONFIG_I2C_I801=m
11)
CONFIG_I2C_DESIGNWARE_CORE=m
CONFIG_I2C_DESIGNWARE_SLAVE=y
CONFIG_I2C_DESIGNWARE_PLATFORM=m
# CONFIG_I2C_DESIGNWARE_AMDPSP is not set
CONFIG_I2C_DESIGNWARE_BAYTRAIL=y
CONFIG_I2C_DESIGNWARE_PCI=m
12)
CONFIG_PINCTRL_BAYTRAIL=y
CONFIG_PINCTRL_CHERRYVIEW=y
CONFIG_PINCTRL_LYNXPOINT=m
CONFIG_PINCTRL_INTEL=y
CONFIG_PINCTRL_ALDERLAKE=m
CONFIG_PINCTRL_BROXTON=m
CONFIG_PINCTRL_CANNONLAKE=m
CONFIG_PINCTRL_CEDARFORK=m
CONFIG_PINCTRL_DENVERTON=m
CONFIG_PINCTRL_ELKHARTLAKE=m
CONFIG_PINCTRL_EMMITSBURG=m
CONFIG_PINCTRL_GEMINILAKE=m
CONFIG_PINCTRL_ICELAKE=m
CONFIG_PINCTRL_JASPERLAKE=m
CONFIG_PINCTRL_LAKEFIELD=m
CONFIG_PINCTRL_LEWISBURG=m
CONFIG_PINCTRL_METEORLAKE=m
CONFIG_PINCTRL_SUNRISEPOINT=m
CONFIG_PINCTRL_TIGERLAKE=m
13)
CONFIG_GPIO_ICH=m
14)
CONFIG_LPC_ICH=m
15)
CONFIG_INTEL_IDMA64=m |
1. YES ... 
2. I recommend to enable it with 1
3. This has nothing to do with our problem, but you really should change it, because it cause a huge performance impact if you dont have an user damon (CPU is running at lowest frequency all the time). Recommendation for a notebook is "powersafe", but "schedutil" is also fine.
4. This is an example of a line I am missing ... have you deleted it manually ? (but it is not important)
5. Enable it
6. HERE it could be the reason of your problem. Enable all these with <*>
7. Enable it
8. I'm surprised you don't need this (behind is SATA) ... but I dont have your "lsmod"
9. Enable all these with <*>
10. Enable all these with <*>
11. If you have a designware controller ("lsmod"!) enable it static with <*>
12. You will need one of them with <*>
13. HERE it could be the reason of your problem. Enable all these with <*>
14. HERE it could be the reason of your problem. Enable all these with <*>
15. If it is in your lsmod-list then enable it static with <*> |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54782
Location: 56N 3W
|
| Posted: Tue Aug 29, 2023 11:36 am Post subject: |
|
|
dk0393,
eMMC and SD cards are both asynchronously detected by the kernel.
Coldplug (at boot) fails as its been and gone by the time that th SD card shows up.
The eMMC has the same problem but the kernel has special provisions to deal with that as root in there.
Logs are good for two things.
1. The things that they show, that should be there.
2. The things that they don't show that should be there.
We can't see the second group from log fragments, so full logs posted with wgetpaste are much preferred.
As you want to used the SD card for additional storage, it needs to be detected before localmount runs.
That may require some extra jiggery pokery in an init script.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Tue Aug 29, 2023 11:41 am Post subject: |
|
|
| szatox wrote: | | Quote: | | SD Card is not discovered on boot, ok if re-inserted. |
"Works after hotplug but not after reboot" looks like udev to me.
You can trigger udev events manually with udevadm trigger -c add|change. Check out the manual for more options (and for ways to limit the scope of trigger if you need it)
Also, I wonder if there is any easy way to get a debug log from udev after running a trigger, it could help with fixing the problem with X not working after reboot. |
Well, actually I was playing around with udev right before seing this comment Narrowed it with -y option, but in vain. |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Tue Aug 29, 2023 12:04 pm Post subject: |
|
|
pietinger, Many thanks, I will revise the config.
I actually never edited it manually, only with oldconfig/menuconfig. Some selections might look odd just because I'm rather new to this and only built-in the things I was positively sure about (and in some cases forgot to remove excessive modules after gaining some knowledge).
as to lsmod, sorry, I completely forgot about it Here it is.
NeddySeagoon, do you happen to know if there are any docs on those kernel provisions for eMMC? Would be glad to read if any...
Sorry, my bad with logs, nex time I'll post them at once
As to detection of sdcard before localmount, if I'm not mistaken, dracut permits to add some custom script as a pre-mount module, so this should do the trick. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Tue Aug 29, 2023 12:38 pm Post subject: |
|
|
| dk0393 wrote: | | as to lsmod, sorry, I completely forgot about it |
Please don't worry; this is no problem. Since I don't know how your sd device is tethered in your Asus chromebook cx1400, I recommended doing as much built-in as possible (it may not be necessary, but it won't hurt either; I am also unsure if you have a designware PCI controller, but if yes, maybe it is connected to it?). That's what else I saw (CRC is required by each block device):
| Code: | crct10dif_pclmul 16384 1
crc32_pclmul 16384 0
crc32c_intel 24576 0
sha512_ssse3 49152 0
lpc_ich 28672 0
i2c_i801 36864 0
i2c_smbus 20480 1 i2c_i801
idma64 20480 0
i2c_designware_platform 16384 0
i2c_designware_core 32768 1 i2c_designware_platform
pinctrl_broxton 36864 1 |
Maybe try to enable these also static:
| Code: | # DesignWare PCI Core Support
#
# CONFIG_PCIE_DW_PLAT_HOST is not set
# CONFIG_PCI_MESON is not set
# end of DesignWare PCI Core Support
CONFIG_CRYPTO_SHA512_SSSE3=m
CONFIG_CRYPTO_CRC32C_INTEL=m
CONFIG_CRYPTO_CRC32_PCLMUL=m
CONFIG_CRYPTO_CRCT10DIF_PCLMUL=m |
If this all doesnt help I would need a new "dmesg" from systemboot-and-reinsert-sdcard after systemboot. What you can also do (if you want):
- Booting an UbuntuLiveCD and check if you have the same problem.
If it works there, then do also a "lsmod" and compare it to your module configuration (...and later with udev settings). |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54782
Location: 56N 3W
|
| Posted: Tue Aug 29, 2023 3:42 pm Post subject: |
|
|
dk0393,
You probably know dracut better that me. I build my kernels by hand and when I need an initrd, I make that by hand too.
I don't have any autoblackmagic to trip me up here.
Even my car has a manual choke :)
I don't know of any useful MMC docs. The API is online but its probably not what you wat.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
logrusx
Advocate
Joined: 22 Feb 2018
Posts: 2656
|
| Posted: Tue Aug 29, 2023 4:31 pm Post subject: |
|
|
It might be a design issue of the reader itself. Why don't you try with a live distribution, for example Ubuntu. If it works there, it's a matter of settings. If not - it's still unknown.
Best Regards,
Georgi |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54782
Location: 56N 3W
|
| Posted: Tue Aug 29, 2023 5:08 pm Post subject: |
|
|
dk0393,
Lets try some analysis and look for supporting or contradictory evidence.
Either is useful. This is about establishing what happens, not about fixing it yet.
You boot from emmc with the SD card plugged in.
When the dust settles, you should have two different /dev/mmc* if all is well.
If one is missing totally, is it the SD card?
As a half way house, the SD card may get detected after coldplug has done its thing but before hotplug is ready.
In this case, the /dev entry will still by created by the kernels DEVTMPFS but the event that udev is supposed to handle will be dropped on the floor an the /dev/ node for the SD card will be present with the wrong permissions.
Is the SD card /dev node missing entirely or does it have the wrong permissions?
Do you get the same result every time?
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Tue Aug 29, 2023 5:59 pm Post subject: |
|
|
Neddy,
AFAIK a sd card works with encryption. If this is built in the kernel as module, then the kernel is not able to access it (this is the reason why I try to configure these things static into his kernel; of course I dont know what his initramfs does, but this doesnt matter if we have all necessary modules in the kernel built-in static). Maybe we should wait for a feedback from PO ? (He has a lot to configure) |
|
| Back to top |
|
|
grknight
Retired Dev
Joined: 20 Feb 2015
Posts: 1994
|
| Posted: Tue Aug 29, 2023 6:46 pm Post subject: |
|
|
| pietinger wrote: | | AFAIK a sd card works with encryption. |
An SD Card can be encrypted in the normal ways but it is just a flash media otherwise (the default). The origins of "Secure Digital" is MMC + DRM (originally for MP3s) and some form factor changes. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Tue Aug 29, 2023 6:51 pm Post subject: |
|
|
| grknight wrote: | | An SD Card can be encrypted in the normal ways but it is just a flash media otherwise (the default). The origins of "Secure Digital" is MMC + DRM (originally for MP3s) and some form factor changes. |
Eew, thanks a lot, grknight. I have thought it is always a "secure digital" (I knew this term) and didnt knew that it can be a normal flash also.  |
|
| Back to top |
|
|
szatox
Advocate
Joined: 27 Aug 2013
Posts: 3481
|
| Posted: Tue Aug 29, 2023 8:23 pm Post subject: |
|
|
Doesn't the "secure" part of secure digital come from write protection mechanical switch? I doubt those things have enough electronics inside to implement any kind of copy restriction.
Anyway...
| Quote: | | Well, actually I was playing around with udev right before seing this comment Narrowed it with -y option, but in vain. |
so I guess it didn't work right of the bat. Well, narrowing it down is not necessary, by default it triggers a system-wide event which should result in udev processing all the rules for the selected type.
I've been using this hack for enabling my GSM modem, it often required several shots to properly take effect, so I started doing 4 times add + 4 times change after reboot (with like 300 ms sleep after each trigger) before launching wvdial. 'Twas Ugly, but worked well enough.
Regarding logs, dmesg might contain some useful information.
Try to boot with that card in, then hotplug it again, and see what's going on |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Wed Aug 30, 2023 11:23 am Post subject: |
|
|
Hi all!
pietinger, I recompiled the kernel with the last recommended parameters (crypto), but that didn't change anything (built-in not only those that you mentioned, but all of them, as it was said, that would not hurt). Dmesg output after boot and after replugging.
pietinger, NeddySeagoon, logrusx, Unfortunately, I don't have ubuntu live cd by hand, but I used gentoo installation media for the same purpose. Had no luck with both gui live cd and minimal one.
| NeddySeagoon wrote: | ...
You boot from emmc with the SD card plugged in.
When the dust settles, you should have two different /dev/mmc* if all is well.
If one is missing totally, is it the SD card?
As a half way house, the SD card may get detected after coldplug has done its thing but before hotplug is ready.
In this case, the /dev entry will still by created by the kernels DEVTMPFS but the event that udev is supposed to handle will be dropped on the floor an the /dev/ node for the SD card will be present with the wrong permissions.
Is the SD card /dev node missing entirely or does it have the wrong permissions?
Do you get the same result every time? |
Each time the computer boots up /dev/mmcblk* for SD card is missing, and internal eMMC is always there, this repeats every boot.
There never were any issues with access rights, it is simply missing. |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Wed Aug 30, 2023 12:38 pm Post subject: |
|
|
We have a change ... Look at "mmcblk1" and "mmcblk0" ... Before (in first post) your MMC card was "mmcblk0: p1 p2 p3" and now it is the second ... this means kernel has given your SDSX card "mmcblk0: p1" ... and therefore kernel has recognized it ... now we would need to know what your intramfs is doing ... or your systemd (together with udev).
For me kernel is okay now ... and ... you dont have an designware PCI controller (so you might disable it).
| Code: | [ 0.583646] mmc1: new HS400 MMC card at address 0001
[ 0.584243] mmcblk1: mmc1:0001 TX2964 58.3 GiB
[ 0.587117] mmcblk1: p1 p2 p3
[ 0.587916] mmcblk1boot0: mmc1:0001 TX2964 4.00 MiB
[ 0.588909] mmcblk1boot1: mmc1:0001 TX2964 4.00 MiB
[ 0.589686] mmcblk1rpmb: mmc1:0001 TX2964 4.00 MiB, chardev (243:0
[...]
[ 192.853528] mmc0: new ultra high speed SDR104 SDXC card at address 59b4
[ 192.861981] mmcblk0: mmc0:59b4 SDU1 58.6 GiB
[ 192.866488] mmcblk0: p1 |
Unfortunately i don't know much about systemd ... but we have experts for this ... 
Sidenote to kernel:
This is okay - dont worry:
| Code: | [ 0.300223] pinctrl core: failed to create debugfs directory for INT3452:00
[ 0.470455] intel-lpss 0000:00:16.0: Failed to create debugfs entries |
(because you dont have enabled debugfs - This is quite fine - I also dont have it because => Security risk !)
Here I was surprised:
| Code: | | [ 0.364016] AVX or AES-NI instructions are not detected. |
... is really such a crippled Intel CPU ? (You surly know now what you dont need; but it doesnt hurt if you dont disable it; every module in kernel which will not work, will not be used and cleaned out from memory later)
If you have time you might disbale this - it is only for XEON CPUs:
| Code: | | [ 0.466217] idxd driver failed to load without MOVDIR64B. |
=> https://www.kernelconfig.io/config_intel_idxd
The rest of your dmesg is looking really good to me.  |
|
| Back to top |
|
|
pietinger
Moderator
Joined: 17 Oct 2006
Posts: 5347
Location: Bavaria
|
| Posted: Wed Aug 30, 2023 12:49 pm Post subject: |
|
|
P.P.S.
If you have really much time and want to investigate further you could just set these static (and let all other as <M>odule):
| Code: | CONFIG_BLK_DEV_SD=m
CONFIG_BLK_DEV_SR=m
CONFIG_INTEL_IDMA64=m
CONFIG_PINCTRL_BROXTON=m |
(I believe these are the responsible modules) |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54782
Location: 56N 3W
|
| Posted: Wed Aug 30, 2023 1:33 pm Post subject: |
|
|
dk0393,
Can we see your current full kernel .config and dmesg please?
I'm mostly interested in it not working.
wgetpaste is your friend.
Your output would be good too.
My 2016 Acer R13 Chromebook has a very small eMMC card, which still holds Chromium, then boots Gentoo from the SD card.
Getting rid of Chromium is risky, so that's the least worst option for me.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Wed Aug 30, 2023 4:56 pm Post subject: |
|
|
| pietinger wrote: | | We have a change ... Look at "mmcblk1" and "mmcblk0" |
Noticed that, but isn't it related to the order of enumeration and naming of my pci devices? In previous dmesg outputs we had:
| Code: | [ 0.565909] mmc0: new HS400 MMC card at address 0001
...
[ 6.127261] mmc1: SDHCI controller on PCI [0000:00:1b.0] using ADMA 64-bit
|
and now it is:
| Code: | [ 0.583646] mmc1: new HS400 MMC card at address 0001
...
[ 0.497449] mmc0: SDHCI controller on PCI [0000:00:1b.0] using ADMA 64-bit
|
| pietinger wrote: | | you dont have an designware PCI controller (so you might disable it) |
For some reason my ELAN touchpad refuses to work without this driver (tried different solutions, and only this one worked)
| pietinger wrote: | | is really such a crippled Intel CPU ? |
Celeron n3350, not too capable one
As to non-working modules, will disable them later.
Already enabled as built-in
CONFIG_BLK_DEV_SD
CONFIG_BLK_DEV_SR
CONFIG_INTEL_IDMA64
CONFIG_PINCTRL_BROXTON
| NeddySeagoon wrote: | | Can we see your current full kernel .config and dmesg please? |
Sure dmesg available in this post, it atually hasn't changed... and kernel config is available here
Also, here is lspci -nnk |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54782
Location: 56N 3W
|
| Posted: Wed Aug 30, 2023 5:27 pm Post subject: |
|
|
dk0393,
From dmesg ...
| Code: | [ 0.479463] sdhci: Secure Digital Host Controller Interface driver
[ 0.479468] sdhci: Copyright(c) Pierre Ossman
[ 0.479551] sdhci-pci 0000:00:1b.0: SDHCI controller found [8086:5aca] (rev b)
[ 0.481254] sdhci-pci 0000:00:1c.0: SDHCI controller found [8086:5acc] (rev b)
[ 0.482759] mmc0: SDHCI controller on PCI [0000:00:1c.0] using ADMA 64-bit |
The kernel found both your controllers and it looks like the one at PCI [0000:00:1c.0] is detected with something in it.
At this level, we cannot tell if its eMMC on an SD card as the interfaces are identical.
| Code: | [ 0.565909] mmc0: new HS400 MMC card at address 0001
[ 0.566648] mmcblk0: mmc0:0001 TX2964 58.3 GiB
[ 0.569685] mmcblk0: p1 p2 p3
[ 0.570499] mmcblk0boot0: mmc0:0001 TX2964 4.00 MiB
[ 0.571617] mmcblk0boot1: mmc0:0001 TX2964 4.00 MiB
[ 0.572552] mmcblk0rpmb: mmc0:0001 TX2964 4.00 MiB, chardev (243:0) |
mmc0 is your eMMC.
Those boot0, boot1 and rpmb are read only space for Chromium.
| Code: | | [ 0.572671] sdhci-pci 0000:00:1b.0: SDHCI controller found [8086:5aca] (rev b) |
Thats the SD card slot, again.
The only sign of mmc1 is
| Code: | | [ 6.127261] mmc1: SDHCI controller on PCI [0000:00:1b.0] using ADMA 64-bit |
It's like the slot did not detect a SD card inserted.
Your lspci -nnk is good and nothing stands out in you kernel.
Can you cycle the card in the card slot, then post dmesg again (all of it please) that will show what happens when it does not work and when it does.
There may be some hints there.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
dk0393
n00b
Joined: 24 Aug 2023
Posts: 10
|
| Posted: Wed Aug 30, 2023 5:54 pm Post subject: |
|
|
| NeddySeagoon wrote: | dk0393,
...
Can you cycle the card in the card slot, then post dmesg again (all of it please) that will show what happens when it does not work and when it does.
There may be some hints there. |
yep, already available here
The hardware doesn't seem to see the moment of unplugging sd card, but spots it immediately after it is plugged back in. |
|
| Back to top |
|
|
|
Kernel & Hardware
