Discussion:
BIOS vs UEFI
(too old to reply)
muta...@gmail.com
2021-10-20 23:42:05 UTC
Permalink
It seems to me that the BIOS concept of reading the
first sector of the disk into memory and immediately
executing it, from offset 0, is a lot cleaner than the
UEFI concept of requiring a FAT-formatted disk with
subdirectories.

If you're going to go for the latter option of a FAT-formatted
disk, shouldn't it be something internal to the computer, on
a flash drive, that results in loading the first sector from the hard
disk and executing it, but perhaps in PM32 or LM64 instead
of RM16?

Passing the UEFI data structure to the code on the first
sector would seem to be more appropriate than doing
interrupts though.

BFN. Paul.
Branimir Maksimovic
2021-10-21 01:20:05 UTC
Permalink
Post by ***@gmail.com
It seems to me that the BIOS concept of reading the
first sector of the disk into memory and immediately
executing it, from offset 0, is a lot cleaner than the
UEFI concept of requiring a FAT-formatted disk with
subdirectories.
If you're going to go for the latter option of a FAT-formatted
disk, shouldn't it be something internal to the computer, on
a flash drive, that results in loading the first sector from the hard
disk and executing it, but perhaps in PM32 or LM64 instead
of RM16?
Passing the UEFI data structure to the code on the first
sector would seem to be more appropriate than doing
interrupts though.
BFN. Paul.
UEFI sucks as it forces FAT32 partition and Windows executable...
--
7-77-777
Evil Sinner!
with software, you repeat same experiment, expecting different results...
Grant Taylor
2021-10-21 01:52:13 UTC
Permalink
Post by Branimir Maksimovic
UEFI sucks as it forces FAT32 partition and Windows executable...
UEFI does not force Windows executable(s).

Linux executable(s) work perfectly fine with UEFI.
--
Grant. . . .
unix || die
Branimir Maksimovic
2021-10-21 02:50:12 UTC
Permalink
Post by Grant Taylor
Post by Branimir Maksimovic
UEFI sucks as it forces FAT32 partition and Windows executable...
UEFI does not force Windows executable(s).
Linux executable(s) work perfectly fine with UEFI.
Linux executables in PE format :P
--
7-77-777
Evil Sinner!
with software, you repeat same experiment, expecting different results...
Rod Pemberton
2021-10-25 07:01:08 UTC
Permalink
On Thu, 21 Oct 2021 02:50:12 GMT
Post by Branimir Maksimovic
Post by Grant Taylor
Post by Branimir Maksimovic
UEFI sucks as it forces FAT32 partition and Windows executable...
UEFI does not force Windows executable(s).
Linux executable(s) work perfectly fine with UEFI.
Linux executables in PE format :P
I'm sure I posted this link recently, like back in May. The table
shows Linux and Windows booting from UEFI. Windows from FAT32. Linux
from both FAT16 and FAT32. If you scroll the table to the right, it
shows the sector sizes, ESP size, and EFI size. The article is a few
years old though.

https://www.ctrl.blog/entry/esp-size-guide.html


Hey, if you guys have some better UEFI links, I'd like to read them as
I don't have a UEFI system yet, perhaps not for a while either, but I'd
like to slowly get up to date.
--
Donald Trump: No oil rigs off the East coast.
Joe Biden: Windfarms off of all our coasts.
Branimir Maksimovic
2021-10-26 01:09:38 UTC
Permalink
Post by Rod Pemberton
On Thu, 21 Oct 2021 02:50:12 GMT
Post by Branimir Maksimovic
Post by Grant Taylor
Post by Branimir Maksimovic
UEFI sucks as it forces FAT32 partition and Windows executable...
UEFI does not force Windows executable(s).
Linux executable(s) work perfectly fine with UEFI.
Linux executables in PE format :P
I'm sure I posted this link recently, like back in May. The table
shows Linux and Windows booting from UEFI. Windows from FAT32. Linux
from both FAT16 and FAT32. If you scroll the table to the right, it
shows the sector sizes, ESP size, and EFI size. The article is a few
years old though.
https://www.ctrl.blog/entry/esp-size-guide.html
Hey, if you guys have some better UEFI links, I'd like to read them as
I don't have a UEFI system yet
Try it it's dead simple and dumb :P


, perhaps not for a while either, but I'd
Post by Rod Pemberton
like to slowly get up to date.
--
7-77-777
Evil Sinner!
with software, you repeat same experiment, expecting different results...
Scott Lurndal
2021-10-26 13:22:54 UTC
Permalink
Post by Rod Pemberton
Hey, if you guys have some better UEFI links, I'd like to read them as
https://uefi.org/

muta...@gmail.com
2021-10-21 05:19:46 UTC
Permalink
Post by Branimir Maksimovic
Post by ***@gmail.com
It seems to me that the BIOS concept of reading the
first sector of the disk into memory and immediately
executing it, from offset 0, is a lot cleaner than the
UEFI concept of requiring a FAT-formatted disk with
subdirectories.
If you're going to go for the latter option of a FAT-formatted
disk, shouldn't it be something internal to the computer, on
a flash drive, that results in loading the first sector from the hard
disk and executing it, but perhaps in PM32 or LM64 instead
of RM16?
Passing the UEFI data structure to the code on the first
sector would seem to be more appropriate than doing
interrupts though.
BFN. Paul.
UEFI sucks as it forces FAT32 partition and Windows executable...
Isn't that what I just said?

But is there anything wrong with the UEFI data structure
being passed to code executing from sector 0, same as
traditional BIOS?

BFN. Paul.
Branimir Maksimovic
2021-10-21 21:05:10 UTC
Permalink
Post by ***@gmail.com
Post by Branimir Maksimovic
Post by ***@gmail.com
It seems to me that the BIOS concept of reading the
first sector of the disk into memory and immediately
executing it, from offset 0, is a lot cleaner than the
UEFI concept of requiring a FAT-formatted disk with
subdirectories.
If you're going to go for the latter option of a FAT-formatted
disk, shouldn't it be something internal to the computer, on
a flash drive, that results in loading the first sector from the hard
disk and executing it, but perhaps in PM32 or LM64 instead
of RM16?
Passing the UEFI data structure to the code on the first
sector would seem to be more appropriate than doing
interrupts though.
BFN. Paul.
UEFI sucks as it forces FAT32 partition and Windows executable...
Isn't that what I just said?
But is there anything wrong with the UEFI data structure
being passed to code executing from sector 0, same as
traditional BIOS?
BFN. Paul.
They wanted to eliminate need for bootloader...
nothing wrong,,,
--
7-77-777
Evil Sinner!
with software, you repeat same experiment, expecting different results...
muta...@gmail.com
2021-10-21 21:47:15 UTC
Permalink
Post by Branimir Maksimovic
Post by Branimir Maksimovic
UEFI sucks as it forces FAT32 partition and Windows executable...
They wanted to eliminate need for bootloader...
I have some more information from the osdev discord.

1. It's probably wrong to embed knowledge of even a
partition table in firmware, nevermind an entire file system.

2. The UEFI shell is a viable DOS-like OS in its own right - they
have put an entire OS into firmware and you can write applications
for it if you want. What I haven't found out is whether the shell
supports ANSI escape sequences.

BFN. Paul.
Loading...