View Issue Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0001010 | channel: elrepo/el8 | --elrepo--request-for-enhancement-- | public | 2020-06-16 13:03 | 2020-06-19 12:51 |
Reporter | lowen | Assigned To | pperry | ||
Priority | normal | Severity | feature | Reproducibility | always |
Status | resolved | Resolution | fixed | ||
Summary | 0001010: EL 8 kernels do not support all qla2xxx PCI IDs. | ||||
Description | Upstream EL8's kernel has support for a number of the PCI IDs covered by the qla2xxx driver removed. Requesting a kmod if possible for qla2xxx to add those IDs back. | ||||
Additional Information | PCI IDs impatced are listed under the qla2xxx driver on https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/8/html/considerations_in_adopting_rhel_8/hardware-enablement_considerations-in-adopting-rhel-8 | ||||
Tags | No tags attached. | ||||
|
Will work on it. In the meantime, could you try installing the centosplus kernel (kernel-plus)? It is supposed to have the qla2xxx module. |
|
My bad. Those removed IDs are not in the plus kernel. kernel-ml is the one you may want to try. alias: pci:v00001077d00002289sv*sd*bc*sc*i* alias: pci:v00001077d00002089sv*sd*bc*sc*i* alias: pci:v00001077d00002281sv*sd*bc*sc*i* alias: pci:v00001077d00002081sv*sd*bc*sc*i* alias: pci:v00001077d00002061sv*sd*bc*sc*i* alias: pci:v00001077d00002261sv*sd*bc*sc*i* alias: pci:v00001077d00002271sv*sd*bc*sc*i* alias: pci:v00001077d00002071sv*sd*bc*sc*i* alias: pci:v00001077d00008044sv*sd*bc*sc*i* alias: pci:v00001077d0000F001sv*sd*bc*sc*i* alias: pci:v00001077d00008031sv*sd*bc*sc*i* alias: pci:v00001077d00008021sv*sd*bc*sc*i* alias: pci:v00001077d00008001sv*sd*bc*sc*i* alias: pci:v00001077d00002031sv*sd*bc*sc*i* alias: pci:v00001077d00002532sv*sd*bc*sc*i* alias: pci:v00001077d00005432sv*sd*bc*sc*i* alias: pci:v00001077d00005422sv*sd*bc*sc*i* alias: pci:v00001077d00008432sv*sd*bc*sc*i* alias: pci:v00001077d00002432sv*sd*bc*sc*i* alias: pci:v00001077d00002422sv*sd*bc*sc*i* alias: pci:v00001077d00006322sv*sd*bc*sc*i* alias: pci:v00001077d00006312sv*sd*bc*sc*i* alias: pci:v00001077d00002322sv*sd*bc*sc*i* alias: pci:v00001077d00002312sv*sd*bc*sc*i* alias: pci:v00001077d00002300sv*sd*bc*sc*i* alias: pci:v00001077d00002200sv*sd*bc*sc*i* alias: pci:v00001077d00002100sv*sd*bc*sc*i* |
|
I'll see if I can get you a kmod package built tomorrow. Do you also need a DUD or just the kmod package? |
|
I only need a kmod, since I don't currently boot from fibre-channel. Others who might want /root on SAN might appreciate a DUD. |
|
And, by the way MANY THANKS! |
|
In addition to restoring the removed pci_id's, do we want to enable the TCM_QLA2XXX Kconfig option? This option is enabled on RHEL7 kernels but is not set on RHEL8. I wonder if it's not set on RHEL8 as part of RH disabling support for older adapters, at which point do we want to enable it when building this module? Or do we want to leave it as not set? Here's the relevant Kconfig: config TCM_QLA2XXX tristate "TCM_QLA2XXX fabric module for QLogic 24xx+ series target mode HBAs" depends on SCSI_QLA_FC && TARGET_CORE depends on LIBFC select BTREE default n ---help--- Say Y here to enable the TCM_QLA2XXX fabric module for QLogic 24xx+ series target mode HBA |
|
There is an upstream bugzilla for that, actually, at https://bugzilla.redhat.com/show_bug.cgi?id=1666377 No technical reason is given. Target mode is an interesting and very cool feature, but I've never used it and it seems to have had issues in the past, so unless someone really wants it I would say that it might be best to leave it off. |
|
Ack - built without TCM_QLA2XXX I have built a kmod for el8.2 and uploaded to the testing repository: kmod-qla2xxx-10.01.00.21.08.2-1.el8_2.elrepo.src.rpm kmod-qla2xxx-10.01.00.21.08.2-1.el8_2.elrepo.x86_64.rpm Please test and feedback here if things work as expected (or not) yum --enablerepo=elrepo-testing install kmod-qla2xxx For info: $ rpm -q kmod-qla2xxx kmod-qla2xxx-10.01.00.21.08.2-1.el8_2.elrepo.x86_64 $ modinfo qla2xxx filename: /lib/modules/4.18.0-193.6.3.el8_2.x86_64/weak-updates/qla2xxx/qla2xxx.ko firmware: ql2500_fw.bin firmware: ql2400_fw.bin firmware: ql2322_fw.bin firmware: ql2300_fw.bin firmware: ql2200_fw.bin firmware: ql2100_fw.bin version: 10.01.00.21.08.2-k license: GPL description: QLogic Fibre Channel HBA Driver author: QLogic Corporation rhelversion: 8.2 srcversion: 5FB38F473597C2C1A333825 alias: pci:v00001077d00002289sv*sd*bc*sc*i* alias: pci:v00001077d00002089sv*sd*bc*sc*i* alias: pci:v00001077d00002281sv*sd*bc*sc*i* alias: pci:v00001077d00002081sv*sd*bc*sc*i* alias: pci:v00001077d00002061sv*sd*bc*sc*i* alias: pci:v00001077d00002261sv*sd*bc*sc*i* alias: pci:v00001077d00002271sv*sd*bc*sc*i* alias: pci:v00001077d00002071sv*sd*bc*sc*i* alias: pci:v00001077d00008044sv*sd*bc*sc*i* alias: pci:v00001077d0000F001sv*sd*bc*sc*i* alias: pci:v00001077d00008031sv*sd*bc*sc*i* alias: pci:v00001077d00008021sv*sd*bc*sc*i* alias: pci:v00001077d00008001sv*sd*bc*sc*i* alias: pci:v00001077d00002031sv*sd*bc*sc*i* alias: pci:v00001077d00002532sv*sd*bc*sc*i* alias: pci:v00001077d00005432sv*sd*bc*sc*i* alias: pci:v00001077d00005422sv*sd*bc*sc*i* alias: pci:v00001077d00008432sv*sd*bc*sc*i* alias: pci:v00001077d00002432sv*sd*bc*sc*i* alias: pci:v00001077d00002422sv*sd*bc*sc*i* alias: pci:v00001077d00006322sv*sd*bc*sc*i* alias: pci:v00001077d00006312sv*sd*bc*sc*i* alias: pci:v00001077d00002322sv*sd*bc*sc*i* alias: pci:v00001077d00002312sv*sd*bc*sc*i* alias: pci:v00001077d00002300sv*sd*bc*sc*i* alias: pci:v00001077d00002200sv*sd*bc*sc*i* alias: pci:v00001077d00002100sv*sd*bc*sc*i* depends: nvme-fc,scsi_transport_fc name: qla2xxx vermagic: 4.18.0-193.el8.x86_64 SMP mod_unload modversions sig_id: PKCS#7 signer: ELRepo.org Secure Boot Key sig_key: E9:D4:71:CF:B4:FE:13:6C sig_hashalgo: sha256 |
|
Nice; thanks! Downloaded and installed on the test host; I'll cable up to the SAN and assign LUNs and get back to you. So far though: [root@c8-kvm-pe1950-1 ~]# dmesg|grep qla2xxx [ 12.289183] qla2xxx [0000:00:00.0]-0005: : QLogic Fibre Channel HBA Driver: 10.01.00.21.08.2-k. [ 12.289389] qla2xxx [0000:0a:00.0]-001d: : Found an ISP2432 irq 16 iobase 0x00000000fd227a46. [ 12.346462] qla2xxx [0000:0a:00.0]-ffff:3: FC4 priority set to FCP [ 12.746163] scsi host3: qla2xxx [ 12.748055] qla2xxx [0000:0a:00.0]-00fb:3: QLogic QLE2462 - PCI-Express Dual Channel 4Gb Fibre Channel HBA. [ 12.748063] qla2xxx [0000:0a:00.0]-00fc:3: ISP2432: PCIe (2.5GT/s x4) @ 0000:0a:00.0 hdma+ host#=3 fw=8.07.00 (9496). [ 12.748260] qla2xxx [0000:0a:00.1]-001d: : Found an ISP2432 irq 17 iobase 0x0000000015bd8bc5. [ 12.805294] qla2xxx [0000:0a:00.1]-ffff:4: FC4 priority set to FCP [ 13.069166] scsi host4: qla2xxx [ 13.071019] qla2xxx [0000:0a:00.1]-00fb:4: QLogic QLE2462 - PCI-Express Dual Channel 4Gb Fibre Channel HBA. [ 13.071027] qla2xxx [0000:0a:00.1]-00fc:4: ISP2432: PCIe (2.5GT/s x4) @ 0000:0a:00.1 hdma+ host#=4 fw=8.07.00 (9496). [ 13.071229] qla2xxx [0000:0c:00.0]-001d: : Found an ISP2432 irq 16 iobase 0x000000001438ec72. [ 13.128202] qla2xxx [0000:0c:00.0]-ffff:5: FC4 priority set to FCP [ 13.392158] scsi host5: qla2xxx [ 13.393988] qla2xxx [0000:0c:00.0]-00fb:5: QLogic QLE2462 - PCI-Express Dual Channel 4Gb Fibre Channel HBA. [ 13.393995] qla2xxx [0000:0c:00.0]-00fc:5: ISP2432: PCIe (2.5GT/s x4) @ 0000:0c:00.0 hdma+ host#=5 fw=8.07.00 (9496). [ 13.394156] qla2xxx [0000:0c:00.1]-001d: : Found an ISP2432 irq 17 iobase 0x000000002178e8fa. [ 13.451107] qla2xxx [0000:0c:00.1]-ffff:6: FC4 priority set to FCP [ 13.715154] scsi host6: qla2xxx [ 13.717046] qla2xxx [0000:0c:00.1]-00fb:6: QLogic QLE2462 - PCI-Express Dual Channel 4Gb Fibre Channel HBA. [ 13.717053] qla2xxx [0000:0c:00.1]-00fc:6: ISP2432: PCIe (2.5GT/s x4) @ 0000:0c:00.1 hdma+ host#=6 fw=8.07.00 (9496). [ 33.951050] qla2xxx [0000:0a:00.0]-8038:3: Cable is unplugged... [ 34.271048] qla2xxx [0000:0a:00.1]-8038:4: Cable is unplugged... [ 34.591036] qla2xxx [0000:0c:00.0]-8038:5: Cable is unplugged... [ 34.911040] qla2xxx [0000:0c:00.1]-8038:6: Cable is unplugged... [root@c8-kvm-pe1950-1 ~]# So far so good.... |
|
Great. For reference, I have based the driver off the RHEL 8.2 kernel source code, rather than doing a backport from a kernel.org vanilla kernel. My reasoning being that RH (very) actively maintain this driver (ex the excluded dev_ids) so probably better to base off that in this case. We will almost certainly need to rebuild the driver for each rhel point release, and often these rebuilds break backward compatibility. As above, we denote the build kernel in the release tag (e.g, el8_2) for reference. |
|
Ok, cabled up the box to the SAN, set up a LUN for it, and ran bonnie++ on the LUN. Looks like it's working! |
|
Brilliant. Moving packages from testing to the main repo now. Please feel free to open a new bug if you have any issues. We will likely need to rebuild the driver at point release time whilst RHEL8 is still in development phase. Marking resolved :-) |
Date Modified | Username | Field | Change |
---|---|---|---|
2020-06-16 13:03 | lowen | New Issue | |
2020-06-16 13:03 | lowen | Status | new => assigned |
2020-06-16 13:03 | lowen | Assigned To | => toracat |
2020-06-16 13:21 | toracat | Note Added: 0006958 | |
2020-06-16 13:25 | burakkucat | Project | channel: kernel/el8 => channel: elrepo/el8 |
2020-06-16 13:26 | burakkucat | Category | --kernel--request-for-enhancement-- => --elrepo--request-for-enhancement-- |
2020-06-16 16:59 | pperry | Assigned To | toracat => pperry |
2020-06-16 17:20 | toracat | Note Added: 0006960 | |
2020-06-16 17:40 | pperry | Note Added: 0006962 | |
2020-06-16 20:07 | lowen | Note Added: 0006963 | |
2020-06-16 20:08 | lowen | Note Added: 0006964 | |
2020-06-17 06:07 | pperry | Note Added: 0006965 | |
2020-06-17 06:55 | lowen | Note Added: 0006966 | |
2020-06-17 09:51 | pperry | Note Added: 0006967 | |
2020-06-17 12:31 | lowen | Note Added: 0006968 | |
2020-06-17 13:12 | pperry | Note Added: 0006969 | |
2020-06-19 11:03 | lowen | Note Added: 0006970 | |
2020-06-19 12:51 | pperry | Note Added: 0006971 | |
2020-06-19 12:51 | pperry | Status | assigned => resolved |
2020-06-19 12:51 | pperry | Resolution | open => fixed |