Re: [Qemu-devel] [PULL 08/15] pc-bios/s390-ccw: fix loadparm initialization and int conversion

[Cornelia Huck <cohuck@redhat.com>]

On Tue, 15 May 2018 10:45:02 +0200
Laszlo Ersek <lersek@redhat.com> wrote:

> I've come across this patch in downstream review (although I really have
> zero background in s390x), and Cornelia suggested I might want to repeat
> my comments on the upstream list too:
> 
> On 05/04/18 09:25, Cornelia Huck wrote:
> > From: Collin Walling <walling@linux.ibm.com>
> > 
> > Rename the loadparm char array in main.c to loadparm_str and
> > increased the size by one byte to account for a null termination
> > when converting the loadparm string to an int  via atoui. We
> > also allow the boot menu to be enabled when loadparm is set to
> > an empty string or a series of spaces.
> > 
> > Signed-off-by: Collin Walling <walling@linux.ibm.com>
> > Reported-by: Vasily Gorbik <gor@linux.ibm.com>
> > Reviewed-by: Thomas Huth <thuth@redhat.com>
> > Reviewed-by: Janosch Frank <frankja@linux.ibm.com>
> > Signed-off-by: Thomas Huth <thuth@redhat.com>
> > ---
> >  hw/s390x/ipl.c          |  4 ++++
> >  pc-bios/s390-ccw/main.c | 14 +++++++-------
> >  2 files changed, 11 insertions(+), 7 deletions(-)
> > 
> > diff --git a/hw/s390x/ipl.c b/hw/s390x/ipl.c
> > index fb554ab156..150f6c0582 100644
> > --- a/hw/s390x/ipl.c
> > +++ b/hw/s390x/ipl.c
> > @@ -373,6 +373,10 @@ int s390_ipl_set_loadparm(uint8_t *loadparm)
> >              loadparm[i] = ascii2ebcdic[(uint8_t) lp[i]];
> >          }
> >  
> > +        if (i < 8) {
> > +            memset(loadparm + i, 0x40, 8 - i); /* fill with EBCDIC spaces */
> > +        }
> > +
> >          g_free(lp);
> >          return 0;
> >      }
> > diff --git a/pc-bios/s390-ccw/main.c b/pc-bios/s390-ccw/main.c
> > index 9d9f8cf4d3..26f9adf84a 100644
> > --- a/pc-bios/s390-ccw/main.c
> > +++ b/pc-bios/s390-ccw/main.c
> > @@ -15,11 +15,11 @@
> >  char stack[PAGE_SIZE * 8] __attribute__((__aligned__(PAGE_SIZE)));
> >  static SubChannelId blk_schid = { .one = 1 };
> >  IplParameterBlock iplb __attribute__((__aligned__(PAGE_SIZE)));
> > -static char loadparm[8] = { 0, 0, 0, 0, 0, 0, 0, 0 };
> > +static char loadparm_str[9] = { 0, 0, 0, 0, 0, 0, 0, 0, 0 };
> >  QemuIplParameters qipl;
> >  
> >  #define LOADPARM_PROMPT "PROMPT  "
> > -#define LOADPARM_EMPTY  "........"
> > +#define LOADPARM_EMPTY  "        "
> >  #define BOOT_MENU_FLAG_MASK (QIPL_FLAG_BM_OPTS_CMD | QIPL_FLAG_BM_OPTS_ZIPL)
> >  
> >  /*
> > @@ -45,7 +45,7 @@ void panic(const char *string)
> >  
> >  unsigned int get_loadparm_index(void)
> >  {
> > -    return atoui(loadparm);
> > +    return atoui(loadparm_str);
> >  }
> >  
> >  static bool find_dev(Schib *schib, int dev_no)
> > @@ -80,13 +80,13 @@ static bool find_dev(Schib *schib, int dev_no)
> >  
> >  static void menu_setup(void)
> >  {
> > -    if (memcmp(loadparm, LOADPARM_PROMPT, 8) == 0) {
> > +    if (memcmp(loadparm_str, LOADPARM_PROMPT, 8) == 0) {
> >          menu_set_parms(QIPL_FLAG_BM_OPTS_CMD, 0);
> >          return;
> >      }
> >  
> >      /* If loadparm was set to any other value, then do not enable menu */
> > -    if (memcmp(loadparm, LOADPARM_EMPTY, 8) != 0) {
> > +    if (memcmp(loadparm_str, LOADPARM_EMPTY, 8) != 0) {
> >          return;
> >      }
> >  
> > @@ -116,8 +116,8 @@ static void virtio_setup(void)
> >       */
> >      enable_mss_facility();
> >  
> > -    sclp_get_loadparm_ascii(loadparm);
> > -    memcpy(ldp + 10, loadparm, 8);
> > +    sclp_get_loadparm_ascii(loadparm_str);
> > +    memcpy(ldp + 10, loadparm_str, 8);
> >      sclp_print(ldp);
> >  
> >      memcpy(&qipl, early_qipl, sizeof(QemuIplParameters));
> >   
> 
> So here I said, "There are a bunch of naked 8 constants; those should
> all be (sizeof loadparm_str - 1), I think. Should be a separate patch,
> of course."

Alternatively, as loadparm size is architectured IIRC and unlikely to
change, do

#define LOADPARM_LEN 8
char loadparm_str[LOADPARM_LEN + 1];

and compare against LOADPARM_LEN... opinions?