Skip to content

Commit

Permalink
Update XTeic.adoc
Browse files Browse the repository at this point in the history
  • Loading branch information
jnk0le committed Oct 2, 2024
1 parent 441c48d commit 2bb4fcc
Showing 1 changed file with 7 additions and 8 deletions.
15 changes: 7 additions & 8 deletions XTeic.adoc
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@

= XTeic
Jan Oleksiewicz <jnk0le@hotmail.com>
:appversion: 0.36.1
:appversion: 0.36.2
:toc:
:toclevels: 5
:sectnums:
Expand Down Expand Up @@ -142,12 +142,11 @@ NOTE: ofiicial CLINT is called ACLINT but doesn't differ much from CLINT in sifi

==== generic riscv interrupts as described in "privileged" volume II

Very often refered to as CLINT. e.g. <<sififeintcookbook>> claims that volume II
stuff is CLINT exclusive.
Very often refered to as CLINT. e.g. <<sififeintcookbook>>.

has vectored mode which simply jumps to the position in vector table.
has optional vectored mode which simply jumps to the position in vector table.

Doesn't provide any nesting other than privilege levels
Doesn't provide any nesting other than privilege levels or a complex boilerplate code to disable reatking active interrupts.
Registers and CSR state (`fcsr` etc.) have to be pushed by software before use

==== PLIC/AIA
Expand Down Expand Up @@ -319,7 +318,7 @@ development of such (E)ABI.

Zcmp[e] was also prepared for such fragmentation by reserving first 4 points in rlist for EABI,
so the cores can implement UABI and EABI push/pop instructions at the same time.
Those 4 points are of course supposed to handle 20 caller saved regs of EABI (probably with some
Those 4 points are, of course, supposed to handle 20 caller saved regs of EABI (probably with some
reuse of few higher points).

It will also make the processors capable of stacking 2 registers per cycle, underutilized
Expand Down Expand Up @@ -358,8 +357,8 @@ Recurrig thing in RVE ABI proposals.
The idea is to allow compilers and software vendors to provide a single
set of precompiled libraries for RVI and RVE ABIs.

The issue with this approach is that the code arbitrarily compiled for RVE
might turn out to be less efficient than RVI one. It also limits the capabilities
The issue with this approach is that the code arbitrarily compiled for RVE is likely
to turn out to be less efficient than RVI one. It also limits the capabilities
of RVI ABI like trading off argument registers for temporary/saved ones.

==== one universal standard for everyone use cases
Expand Down

0 comments on commit 2bb4fcc

Please sign in to comment.