Re: [PATCH v5] xen/riscv: dump GPRs and CSRs on unexpected traps

[Oleksii Kurochko <oleksii.kurochko@xxxxxxxxx>]

On 1/28/26 4:17 PM, Jan Beulich wrote:
> On 28.01.2026 13:53, Oleksii Kurochko wrote:
> > --- a/xen/arch/riscv/traps.c
> > +++ b/xen/arch/riscv/traps.c
> > @@ -99,11 +99,65 @@ static const char *decode_cause(unsigned long cause)
> >       return decode_trap_cause(cause);
> >   }
> >   
> > -static void do_unexpected_trap(const struct cpu_user_regs *regs)
> > +static void dump_general_regs(const struct cpu_user_regs *regs)
> >   {
> > -    unsigned long cause = csr_read(CSR_SCAUSE);
> > +#define X(regs, name, delim) \
> > +    printk("%-4s: %016lx" delim, #name, (regs)->name)
> > +
> > +    X(regs, ra, " "); X(regs, sp, "\n");
> > +    X(regs, gp, " "); X(regs, tp, "\n");
> > +    X(regs, t0, " "); X(regs, t1, "\n");
> > +    X(regs, t2, " "); X(regs, s0, "\n");
> > +    X(regs, s1, " "); X(regs, a0, "\n");
> > +    X(regs, a1, " "); X(regs, a2, "\n");
> > +    X(regs, a3, " "); X(regs, a4, "\n");
> > +    X(regs, a5, " "); X(regs, a6, "\n");
> > +    X(regs, a7, " "); X(regs, s2, "\n");
> > +    X(regs, s3, " "); X(regs, s4, "\n");
> > +    X(regs, s5, " "); X(regs, s6, "\n");
> > +    X(regs, s7, " "); X(regs, s8, "\n");
> > +    X(regs, s9, " "); X(regs, s10, "\n");
> > +    X(regs, s11, " "); X(regs, t3, "\n");
> > +    X(regs, t4, " "); X(regs, t5, "\n");
> > +    X(regs, t6, " ");
> DYM "\n" here?
Oh, right, it should be "\n".

> > +#undef X
> > +}
> > +
> > +static void dump_csrs(const char *ctx)
> > +{
> > +#define X(name, csr, fmt, ...) \
> > +    v = csr_read(csr); \
> > +    printk("%-10s: %016lx" fmt, #name, v, ##__VA_ARGS__)
> > +
> > +    unsigned long v;
> > +
> > +    X(htval, CSR_HTVAL, " ");  X(htinst, CSR_HTINST, "\n");
> > +    X(hedeleg, CSR_HEDELEG, " "); X(hideleg, CSR_HIDELEG, "\n");
> > +    X(hstatus, CSR_HSTATUS, " [%s%s%s%s%s%s ]\n",
> > +      (v & HSTATUS_VTSR) ? " VTSR" : "",
> > +      (v & HSTATUS_VTVM) ? " VTVM" : "",
> > +      (v & HSTATUS_HU)   ? " HU"   : "",
> > +      (v & HSTATUS_SPVP) ? " SPVP" : "",
> > +      (v & HSTATUS_SPV)  ? " SPV"  : "",
> > +      (v & HSTATUS_GVA)  ? " GVA"  : "");
> > +    X(hgatp, CSR_HGATP, "\n");
> > +    X(hstateen0, CSR_HSTATEEN0, "\n");
> > +    X(stvec, CSR_STVEC, " "); X(vstvec, CSR_VSTVEC, "\n");
> > +    X(sepc, CSR_SEPC, " "); X(vsepc, CSR_VSEPC, "\n");
> > +    X(stval, CSR_STVAL, " "); X(vstval, CSR_VSTVAL, "\n");
> > +    X(status, CSR_SSTATUS, " "); X(vsstatus, CSR_VSSTATUS, "\n");
> > +    X(satp, CSR_SATP, "\n");
> > +    X(scause, CSR_SCAUSE, " %s[%s]\n", ctx, decode_cause(v));
> For it, in particular the "ctx" string, to stand out, perhaps this wants moving
> first in the function?
I would be grateful for this.

> With the adjustments (happy to carry out while committing, so long as you 
> agree):
> Acked-by: Jan Beulich <jbeulich@xxxxxxxx>
Thanks a lot.

~ Oleksii