RE: Help on TCG asserts

[Sai Pavan Boddu]

Hi Richard,

> -----Original Message-----
> From: Richard Henderson <richard.henderson@linaro.org>
> Sent: Wednesday, October 28, 2020 8:16 PM
> To: Sai Pavan Boddu <saipava@xilinx.com>; QEMU Developers <qemu-
> devel@nongnu.org>
> Cc: Peter Maydell <peter.maydell@linaro.org>; Philippe Mathieu-Daudé
> <f4bug@amsat.org>
> Subject: Re: Help on TCG asserts
> 
> On 10/27/20 9:15 PM, Sai Pavan Boddu wrote:
> > 336 static void set_jmp_reset_offset(TCGContext *s, int which)
> >              337 {
> >              338     size_t off = tcg_current_code_size(s);
> >              339     s->tb_jmp_reset_offset[which] = off;
> >              340     /* Make sure that we didn't overflow the stored
> > offset.  */
> >       >>> 341     assert(s->tb_jmp_reset_offset[which] == off);
> >              342 }
> >
> > Can anyone give some color on what to look out for. As I see, tcg
> > seems to have some restriction of code size it could handle, but I’m
> > confused how to control that in qemu.
> 
> It does mean that the generated code block is too large, and it needs to be
> split.  I'm surprised you hit this though, since it means that something else 
> has
> gone wrong beforehand.
> 
> Is there a test case you can share?
[Sai Pavan Boddu] I am working to get a simple version of the test, as 
testsuite is not sharable. 
I have attached the in_asm/out_asm logs, would be nice if can review and point 
anything going weird.

I am seeing this on xlnx-versal machine.

Regards,
Sai Pavan
> 
> 
> r~

qemu.log
Description: qemu.log