|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Minios-devel] [UNIKRAFT PATCH 2/2] plat/kvm: Update VGA console cursor location
Oops, I see i missed the other comments about the actual code, so let me respond to them here. On 07/30/2018 02:23 PM, Simon Kuenzer wrote: Do you know if we better enable the cursor? I am not sure if we can assume a common state in which the boot loader left us.https://wiki.osdev.org/Text_Mode_Cursor#Enabling_the_Cursor_2 My thought was that this code is specific to KVM at the moment, and I assume that QEMU will always drop us into a defined state. But I can add the code to set the cursor. I'll do it for a v2.
Hmm. I'm not a big fan of disabling interrupts for large amounts of code, including several functions. Long strings can take time to print out, and I'm not sure I feel comfortable disabling interrupts for that long If at all, I would suggest doing that with other kinds of locking mechanisms. I'm also wondering, though, whether it shouldn't be the job of whatever's calling coutk. The typical behavior on most other systems is that, if you print from several threads without synchronization, your output will be interleaved, right? + /* Location of the address and data registers is variable and denoted I honestly don't know. I wouldn't expect this to change after it is set once, but I don't know for sure. Supposedly, 0x3b is mostly there for compatibility with older monochrome adapters, so it probably should be 0x3d pretty much always. If we go by the "this is a driver for QEMU VGA only", we can probably scrap it altogether. Otherwise, it's *probably* fine to just check it once in the beginning. _______________________________________________ Minios-devel mailing list Minios-devel@xxxxxxxxxxxxxxxxxxxx https://lists.xenproject.org/mailman/listinfo/minios-devel
|
![]() |
Lists.xenproject.org is hosted with RackSpace, monitoring our |