Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merge remote-tracking branch 'upstream/master' into difftest #31

Merged
merged 223 commits into from
Sep 6, 2024

Conversation

lewislzh
Copy link

@lewislzh lewislzh commented Sep 5, 2024

No description provided.

YenHaoChen and others added 30 commits May 21, 2024 09:51
Update CSR_MCONTROL6_HIT to CSR_MCONTROL6_HIT0

Include CSR_TINFO_VERSION* macros
…d mcontrol6_t::hit

Add mcontrol_common_t::set_hit()
Avoid using private headers, e.g., debug_defines.h, in triggers.h
Based on Spec chapter 3.5
"An MRET or SRET instruction is used to return from a trap in M-mode or
S-mode, respectively. When executing an xRET instruction, if xPP holds
the value y, then ELP is set to the value of xPELP if yLPE is 1;
otherwise, it is set to NO_LP_EXPECTED; xPELP is set to NO_LP_EXPECTED."

The change follow the last statement after semicolon
"xPELP is set to NO_LP_EXPECTED"

Signed-off-by: Chih-Min Chao <[email protected]>
The spec says: "If vl=0, no operation is performed and the destination
register is not updated." in Section 14. Vector Reduction Operations.

The commit proposes setting the variable is_write to false when vl = 0,
which means not logging the write.
…reduction

vector: Not logging write of reduction instructions when vl = 0
…r/vxsat-write

Require vector extension when attempting vxsat writes
When compiled as PIE, executable can be loaded at any memory address.
Lately, OpenSBI switched to such behavior and spike was not able to load
it anymore. This patch add an additional load_offset parameter for
load_elf(). This load_offset value is passed as DRAM_BASE and used only
for ET_DYN elfs.

Signed-off-by: Clément Léger <[email protected]>
…er/et_dyn

add support to load ET_DYN elf
ELP will be zero if Zicfilp is not enabled.
Serialize after setting ELP.  That way, we can hoist the check
outside of the main simulation loop.
…mplify-zicfilp

Avoid checking ELP before every instruction fetch
Implement Debug spec Section 5.7.6. Trigger Control (tcontrol).

This commit lets tcontrol be read-only 0 if number of triggers is 0.
- c.fsdsp need not be listed since cm.push etc. are listed
- mop.r.28/mop.rr.7 don't have corresponding files in riscv/insns/
- the rest are just erroneous
We get better error checking if we list only the more specific
instructions and omit the more general ones (mop.r.N/mop.rr.N).
aswaterman and others added 17 commits August 27, 2024 16:54
It remains true that PTEs can only represent addresses >= 2^56, but there's
no need to impose that constraint on untranslated accesses.
A previous commit removes the effectiveness of MPRV to MXR.
(riscv-software-src#1784)

However, the removal implies the MPRV affects point masking
individually, and the MXR should consider the effective v bit.
pointer masking: Consider effective v bit instead of current v bit
ISA spec says "Setting MXR at HS-level overrides both VS-stage and G-stage execute-only permissions."
pointer masking: Always apply sstatus.MXR regardless of effective V
…pport-larger-addresses

Lift restriction on physical-address size
…r single-width integer multiply-add instructions
…r widening integer multiply-add instructions
…r single-width floating-point fused multiply-add instructions
…r widening floating-point fused multiply-add instructions
vector: disassemble: Let operand ordering be vd, [vrf]s1, vs2 to vector multiply-add instructions
* Revert "Lift restriction on physical-address size"
	This reverts commit 52f045d.
* fix some  merge compile errors
Copy link
Member

@poemonsense poemonsense left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Look good to me if CI passes

@lewislzh lewislzh merged commit 9f65bc1 into difftest Sep 6, 2024
6 checks passed
@cebarobot cebarobot deleted the dbltrp branch September 10, 2024 13:03
lewislzh added a commit that referenced this pull request Dec 3, 2024
Merge remote-tracking branch 'upstream/master' into difftest
lewislzh added a commit that referenced this pull request Dec 3, 2024
Merge remote-tracking branch 'upstream/master' into difftest
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.