Re: [Qemu-devel] [PATCH 4/5] tcg: reorder removal from lists in tb_phys_invalidate

[Sergey Fedorov]

On 17/03/16 18:09, Paolo Bonzini wrote:
> On 17/03/2016 14:46, address@hidden wrote:
> >   void tb_phys_invalidate(TranslationBlock *tb, tb_page_addr_t page_addr)
> >   {
> > -    CPUState *cpu;
> >       PageDesc *p;
> >       unsigned int h, n1;
> > +    tb_page_addr_t pc;
> >       tb_page_addr_t phys_pc;
> >       TranslationBlock *tb1, *tb2;
> >   
> > -    /* remove the TB from the hash list */
> > -    phys_pc = tb->page_addr[0] + (tb->pc & ~TARGET_PAGE_MASK);
> > -    h = tb_phys_hash_func(phys_pc);
> > -    tb_hash_remove(&tcg_ctx.tb_ctx.tb_phys_hash[h], tb);
> > -
> > -    /* remove the TB from the page list */
> > -    if (tb->page_addr[0] != page_addr) {
> > -        p = page_find(tb->page_addr[0] >> TARGET_PAGE_BITS);
> > -        tb_page_remove(&p->first_tb, tb);
> > -        invalidate_page_bitmap(p);
> > -    }
> > -    if (tb->page_addr[1] != -1 && tb->page_addr[1] != page_addr) {
> > -        p = page_find(tb->page_addr[1] >> TARGET_PAGE_BITS);
> > -        tb_page_remove(&p->first_tb, tb);
> > -        invalidate_page_bitmap(p);
> > -    }
> > -
> > -    tcg_ctx.tb_ctx.tb_invalidated_flag = 1;
> > -
> Did you investigate the removal of this setting of tb_invalidated_flag?
>
> My recollection is that it is okay to remove it because at worse it
> would cause a tb_add_jump from an invalidated source to a valid
> destination.  This should be harmless as long as the source has been
> tb_phys_invalidated and not tb_flushed.  But this needs to be checked.

Thanks for pointing that. I should investigate it to make sure, although 
arm32/arm64/x86_64 Linux boots fine as well as the latest Alex's 
kmv-unit-tests pass.

Kind regards,
Sergey