fozzy8900's Posts

fozzy8900's Posts

I have struggled quite a bit to find an eSATA controller that works in my environment: XPS-8900 Linux 2 Addonics HDD enclosures 4 drives in each enclosure Sil3132 chipset based port-mu... See more...
I have struggled quite a bit to find an eSATA controller that works in my environment: XPS-8900 Linux 2 Addonics HDD enclosures 4 drives in each enclosure Sil3132 chipset based port-multipler Drive capacities ranging from 1.5TB to 5TB Here's the result of my experiments: 1. Syba SD-SA2PEX-2E PCIe 1.0a controller. Not detected by the BIOS, despite Syba and Dell Customer Support's assurances that they should be compatible. 2. StarTech PEXESATA2 PCIe 1.1 controller. Not detected by the BIOS, despite StarTech and Dell Customer Support's assurances that they should be compatible. 3. StarTech PEXESAT32 PCIe 2.0 controller, Marvell 88SE9128 chipset. Successfully detected by the BIOS and the OS, but has trouble establishing reliable connection with the enclosures. Only seems to support one port multiplier at a time. 4. Syba SD-PEX40049 PCIe 2.0 controller, ASM106x chipset. Successfully detected by the BIOS and the OS. Quickly establishes connection with the enclosures. Detects all drives on both ports. So far, seems to correctly support large drives, concurrent I/Os to the same port and hot-plug / hot-unplug. So that's the winner for me. The only caveat is the need to configure 2 sets of 4 jumpers to choose which 2 out of the 4 available ports (2 x eSATA, 2 x SATA) are enabled. That's easy enough to do and is only required once, so not a huge hassle in the big scheme of things.
Blah blah blah. I'm sure most people find your completely besides the point yet assertive statements, all caps, flashy colored diagrams, and debauchery of gratuitous acronyms very impressive. Con... See more...
Blah blah blah. I'm sure most people find your completely besides the point yet assertive statements, all caps, flashy colored diagrams, and debauchery of gratuitous acronyms very impressive. Congratulations. You're a RockStar. I on the other hand am an idiot, my setup is hopeless and I should just give up. Well, that was helpful. Thanks.
The UEFI vs CSM debate is irrelevant to this issue. Both provide a boot environment and an API to the boot loader and the Operating System to interact with the hardware. While they do come into the... See more...
The UEFI vs CSM debate is irrelevant to this issue. Both provide a boot environment and an API to the boot loader and the Operating System to interact with the hardware. While they do come into the picture when dealing with boot loader and OS related problems, they do not when dealing with basic discovery of devices connected to the PCIe bus. The discovery process is driven purely by the PCIe specifications and protocols (which are independent of the platform and the BIOS framework) and happens long before UEFI or CSM are invoked to proceed with the later boot stages. Secure Boot is a subset of UEFI and is therefore also irrelevant to this issue. Secure Boot is about establishing a chain of trust ensuring that the software bits loaded during the boot process (UEFI drivers, boot loader, kernel, ...) come from trusted a source and have not been altered since they were signed. Components not involved in this chain like a TV tuner card or a disk controller not connected to a boot drive are unaffected. All the devices on the PCIe bus, the ones relevant to the software bring up (e.g. disk controller of the boot drive) or the ones not relevant (e.g. TV tuner card, ...) need to be discovered and identified _before_ UEFI and Secure Boot can even know what they are. The issue at hand is that they're not even discovered at the PCIe level as reported by the BIOS itself (not the boot loader, not the OS). PCIe has been designed (very carefully and with much effort) to be fully backward compatible, precisely so people can upgrade their systems progressively and incrementally, without having to throw away all their existing hardware when a new component comes into the picture. As a result, by design PCIe 1.x and 2.x cards can be plugged in PCIe 3.0 slots and operate correctly. They just operate at their native 1.x or 2.x speed, not the 3.0 speed. In many cases, PCIe is even forward compatible i.e. a PCIe 2.x or 3.0 card can be plugged in a 1.x or 2.x slot and function correctly. I am willing to acknowledge that some cards may be victims of ambiguities in parts of the early PCIe specs or of poor implementation of these specs. But dismissing an issue involving different PCIe gens, with no further evidence, is like saying that PCIe is backward compatible except when you use an older version. It's absurd. Finally, I have indeed filed a Service Request with Dell. But I thought that someone in the community might have run into a similar issue and might have relevant information on the matter. This is after all the purpose of these forums, to help each other with constructive comments and actual (i.e. technically correct and relevant) information.
For dual boot, I configured the BIOS to boot the Linux partition first. And grub's OS prober should automatically add an entry at the bottom of your grub menu to let you boot Windows, allowing you ... See more...
For dual boot, I configured the BIOS to boot the Linux partition first. And grub's OS prober should automatically add an entry at the bottom of your grub menu to let you boot Windows, allowing you to boot either of them: Something like this in /etc/grub2-efi.cfg: ### BEGIN /etc/grub.d/30_os-prober ### menuentry 'Windows Boot Manager (on /dev/sda1)' --class windows --class os $menuentry_id_option 'osprober-efi-AF97-6351' { insmod part_gpt insmod fat set root='hd0,gpt1' if [ x$feature_platform_search_hint = xy ]; then  search --no-floppy --fs-uuid --set=root --hint-bios=hd0,gpt1 --hint-efi=hd0,gpt1 --hint-baremetal=ahci0,gpt1  AF97-6351 else  search --no-floppy --fs-uuid --set=root AF97-6351 fi chainloader /EFI/Microsoft/Boot/bootmgfw.efi } Booting Linux from the Windows boot loader is harder.
I ran into the same problem initially: the system would just hang early in the installation process. Editing the grub command line to boot the kernel with the following option solved he issue for me:... See more...
I ran into the same problem initially: the system would just hang early in the installation process. Editing the grub command line to boot the kernel with the following option solved he issue for me:    pcie_aspm=off You obviously have to make it persistent after installation if you don't want to have to edit the command line on every boot.
Also the card works just fine in my aging XPS 9000.
Yeah, I already have Secure Boot disabled. That thing only causes trouble. And I also have legacy BIOS ROM enabled. Still, the BIOS sees nothing in slots 3 and 4.
This is not an Operating System issue (Windows, Linux, FreeBSD, or whatever else). The BIOS itself, whose job is to discover the hardware and prep things up for the OS, reports PCIe slots 3 (x4) an... See more...
This is not an Operating System issue (Windows, Linux, FreeBSD, or whatever else). The BIOS itself, whose job is to discover the hardware and prep things up for the OS, reports PCIe slots 3 (x4) and 4 (x16) as unpopulated even when the card is plugged in.
I recently purchased a XPS-8900. It works beautifully except for one very serious issue: the BIOS doesn't detect my eSATA PCI adapter, a Syba 2 Port eSATA II PCIe x 1 Controller (SD-SA2PEX-2E). I n... See more...
I recently purchased a XPS-8900. It works beautifully except for one very serious issue: the BIOS doesn't detect my eSATA PCI adapter, a Syba 2 Port eSATA II PCIe x 1 Controller (SD-SA2PEX-2E). I need a port multiplier compatible eSATA adapter to access my data, about 20TB of storage, hosted in two eSATA disk enclosures. The enclosures are worth hundreds of dollars by themselves so simply replacing them (e.g. with USB  ones) is a non starter. The symptoms are identical to this XPS-8500 issue which was fixed via a BIOS update (A09). I tried upgrading from the factory BIOS (v2.0.1) to the latest one (v2.0.3), but that did not fix the problem. The PCIe adapter is still not discovered by the BIOS. Anyone else experiencing this ?