optimize dp and dpbase doesn't work in asar 1,90

[gizaha]

-------------------- CODE --------------------------
dpbase $2100
optimize dp always
rep #$30
lda #$1234
sta $2100
optimize dp none

compiled:
c230
a9 3412
8d 0021 IT SHOULD BE 85 00 (sta dp)

[yuriks]

Optimization only applies to labels, not direct addresses. This works:

myLabel = $2100

dpbase myLabel
optimize dp always
rep #$30
lda #$1234
sta myLabel
optimize dp none

C2 30 A9 34 12 85 00

[p4plus2]

Correct, this is by design. We only optimize labels as if somebody hard coded an address they are much more likely to be trying to use a specific address for a reason (and this helps with disassemblies quite a bit).

It is strongly encouraged to use labels for ram and other addresses.

Note: In Yurik's example the dpbase comand uses a label, but it's not necessary when setting the base. Only for the opcodes.

[gizaha]

It's ok. Thank you. There are places with multiple writes to $2116 or $214x and this optimization would save space (using same cycles)

…

On Mon, May 18, 2026 at 10:47 PM p4plus2 ***@***.***> wrote: *p4plus2* left a comment (RPGHacker/asar#400) <#400 (comment)> Correct, this is by design. We only optimize labels as if somebody hard coded an address they are much not likely to be trying to use a specific address for a reason (and this helps with disassemblies quite a bit). It is strongly encouraged to use labels for ram and other addresses. Note: In Yurik's example the dpbase comand uses a label, but it's not necessary when setting the base. Only for the opcodes. — Reply to this email directly, view it on GitHub <#400?email_source=notifications&email_token=AKWMIZCOQCTXIIRAWKBUUDL43NSG3A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTINBYGE2DEMRRGIYKM4TFMFZW63VGMF2XI2DPOKSWK5TFNZ2KYZTPN52GK4S7MNWGSY3L#issuecomment-4481422120>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKWMIZAXZCRSVEHZ2FDYWG343NSG3AVCNFSM6AAAAACZDFORSWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DIOBRGQZDEMJSGA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[p4plus2]

It's ok. Thank you.
There are places with multiple writes to $2116 or $214x and this
optimization would save space (using same cycles)
…

So what I would recommend is using MMIO.asm. Its a file I wrote that will let you use labels/structs to reference MMIO. You can find a copy here: https://gist.github.com/p4plus2/2330c04125df42892e90104bd81bf35d

Quick example:

;uses nullptr from wram as fixed byte source
.clear_vram
    ldx #$80
    stx PPU.vram_control
    lda #DMA_settings(PPU.vram_write, !fixed|!two_regs_once)
    sta DMA[0].settings
    stz PPU.vram_address
    stz DMA[0].source_word
    stz DMA[0].source_bank
    stz DMA[0].size  
    
    ldx #$01
    stx CPU.enable_dma
    rts

This code would basically do the sort of optimizations you are looking for automatically. Barring that, if you wanted to use raw addresses, then you would need to self-optimize.