Closed c0d3z3r0 closed 7 years ago
What's the kernel version?
Oh, I'm sorry... 4.6 and 4.7
having the same with 4.8.0-rc-3.
Could you let me know exact model number of your card? I don't have Trescend one.
Yep, it's a Transcend TS32GUSDU1. I don't have anything than Transcend here so I can't tell if it's a vendor specific problem. @jobenvil is your's a Transcend card, too?
@c0d3z3r0 I have a SanDisk Ultra 8GB Micro SDHC-I Class 10 (this is not an XC, but an HC, thus should be UHS-I, or I'm wrong?). On the back is wroten 6111DKCKT0JN
root@hiperborea:~# dmesg| grep SD
[ 3.774760] sdhci-pltfm: SDHCI platform and OF driver helper
[ 3.876054] mmc1: new high speed SDHC card at address aaaa
Yeah, SanDisk Ultra SDHC is UHS-I, too
@jobenvil did you apply [1] and [2] onto 4.8.0-rc-3?
[1] https://github.com/tobetter/linux/commit/206dbdea8b7a9bc672640c164072383424461393 [2] https://github.com/tobetter/linux/commit/12219a31c0aa8b4d2980c088040330d8dd8f00a9
@c0d3z3r0 No, until you point me out, but still doesn't get recogniced. @tobetter are these patches providing some type of speed increment or only differs in the name denomination?
@jobenvil those patches aren't in mainline. could you try 4.8.0-rc3 with them applied so that we can see if your card is working?
Ok, that means it is not a card specific issue
I'm not sure, thus in my banana pro with kernel 4.6.3 is the same as the XU4 and I don't have more UHS-I cards to test with. Maybe someone listening here wants to post if the UHS-I patches works in his XU4?
The patches are XU3/4 specific - you cannot compare the behaviour of XU3/4 with that of banana pi
Aha... the patches do work but only on cold boot... I'll try to fix this and prepare a patch next week or so
The patches are XU3/4 specific - you cannot compare the behaviour of XU3/4 with that of banana pi
I know, but I saw earlier somewhere on my other ARMs "ultra high speed", I don't remember in which and with which kernel.
I tested your xu4-update
offline and worked fine. Here is the dmesg log.
The SD-card becomes recogniced like UHS-I:
[ 1.728730] mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 50000000Hz, actual 50000000HZ div = 0)
[ 1.729602] mmc1: new ultra high speed DDR50 SDHC card at address aaaa
[ 1.730145] mmcblk1: mmc1:aaaa SL08G 7.40 GiB
[ 1.731142] mmcblk1: p1 p2
Thanks for your efforts and sharing XU4 tools. This is a nice improvement.
Could you check dmesg after a reboot, please
Am 06.09.2016 um 19:33 schrieb José Benlloch notifications@github.com:
The patches are XU3/4 specific - you cannot compare the behaviour of XU3/4 with that of banana pi I know, but I saw earlier somewhere on my other ARMs "ultra high speed", I don't remember in which and with which kernel.
I tested your xu4-update ofline and worked fine. Here is the dmesg log.
The SD-card becomes recogniced like UHS-I:
[ 1.728730] mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 50000000Hz, actual 50000000HZ div = 0) [ 1.729602] mmc1: new ultra high speed DDR50 SDHC card at address aaaa [ 1.730145] mmcblk1: mmc1:aaaa SL08G 7.40 GiB [ 1.731142] mmcblk1: p1 p2 Thanks for your efforts and sharing XU4 tools. This is a nice improvement.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
here after shutdown -r now
confirmed your theory:
[ 1.872958] mmc1: new high speed SDHC card at address aaaa
Thank you! That's what I said yesterday. The patches only work on cold boot, not after a reboot.
testing SD-Card with:
iozone -e -I -a -s 100M -r 4k -r 16k -r 32k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
did show same results on both cases: "ultra high speed vs. high speed", but the UHS-I improvement could be present in other field like, power energy.
@jobenvil my SD card has ~16 MB/s in HS vs. ~70 MB/s in UHS-I mode
Looks like the card has to be reset shortly bevore reboot... I added a mdelay(10000) and reattached the sd card in those 10 seconds... after reboot it's detected as UHS
compiled last tobetter 4.8-rc5. When it succesfully boots with USB3 + UHS-I, I tested the uSD-Card and then, the SSD. The UHS-I patch performs brutally good, awesome because the random read and write values!, but somehow uas could not be recogniced. By other side the ioft/linux repository, with same kernel 4.8-rc5 + uas performs really good, but misses this patch. Comparing both branches, It looks like that these patches ([1] 206dbde and [2] 12219a3)
can not be merged on the top. See yourself the results.
Maybe consider to read this [1] and this [2].
@jobenvil I can apply the patches without any problem. USB 3.0 has nothing to do with UHS
@c0d3z3r0 ptaches on tobetter repository are oke, but try to apply UHS-I patch on top of the ioft repository master branch
ah, I'm sorry - I misunderstood that
Closing this since I sold my Odroid. I'm fed up with ARM...
My Transcend UHS-I SDXC card is not detected as UHS-I
dmesg Odroid-XU4: [ 1.869449] mmc1: new high speed SDXC card at address 59b4
dmesg laptop: [21388.964579] mmc0: new ultra high speed SDR104 SDHC card at address 59b4