[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v4 10/19] xen: setup hypercall page for PVH


  • To: Daniel Kiper <daniel.kiper@xxxxxxxxxx>
  • From: Juergen Gross <jgross@xxxxxxxx>
  • Date: Tue, 20 Nov 2018 10:27:35 +0100
  • Autocrypt: addr=jgross@xxxxxxxx; prefer-encrypt=mutual; keydata= xsBNBFOMcBYBCACgGjqjoGvbEouQZw/ToiBg9W98AlM2QHV+iNHsEs7kxWhKMjrioyspZKOB ycWxw3ie3j9uvg9EOB3aN4xiTv4qbnGiTr3oJhkB1gsb6ToJQZ8uxGq2kaV2KL9650I1SJve dYm8Of8Zd621lSmoKOwlNClALZNew72NjJLEzTalU1OdT7/i1TXkH09XSSI8mEQ/ouNcMvIJ NwQpd369y9bfIhWUiVXEK7MlRgUG6MvIj6Y3Am/BBLUVbDa4+gmzDC9ezlZkTZG2t14zWPvx XP3FAp2pkW0xqG7/377qptDmrk42GlSKN4z76ELnLxussxc7I2hx18NUcbP8+uty4bMxABEB AAHNHkp1ZXJnZW4gR3Jvc3MgPGpncm9zc0BzdXNlLmRlPsLAeQQTAQIAIwUCU4xw6wIbAwcL CQgHAwIBBhUIAgkKCwQWAgMBAh4BAheAAAoJELDendYovxMvi4UH/Ri+OXlObzqMANruTd4N zmVBAZgx1VW6jLc8JZjQuJPSsd/a+bNr3BZeLV6lu4Pf1Yl2Log129EX1KWYiFFvPbIiq5M5 kOXTO8Eas4CaScCvAZ9jCMQCgK3pFqYgirwTgfwnPtxFxO/F3ZcS8jovza5khkSKL9JGq8Nk czDTruQ/oy0WUHdUr9uwEfiD9yPFOGqp4S6cISuzBMvaAiC5YGdUGXuPZKXLpnGSjkZswUzY d9BVSitRL5ldsQCg6GhDoEAeIhUC4SQnT9SOWkoDOSFRXZ+7+WIBGLiWMd+yKDdRG5RyP/8f 3tgGiB6cyuYfPDRGsELGjUaTUq3H2xZgIPfOwE0EU4xwFgEIAMsx+gDjgzAY4H1hPVXgoLK8 B93sTQFN9oC6tsb46VpxyLPfJ3T1A6Z6MVkLoCejKTJ3K9MUsBZhxIJ0hIyvzwI6aYJsnOew cCiCN7FeKJ/oA1RSUemPGUcIJwQuZlTOiY0OcQ5PFkV5YxMUX1F/aTYXROXgTmSaw0aC1Jpo w7Ss1mg4SIP/tR88/d1+HwkJDVW1RSxC1PWzGizwRv8eauImGdpNnseneO2BNWRXTJumAWDD pYxpGSsGHXuZXTPZqOOZpsHtInFyi5KRHSFyk2Xigzvh3b9WqhbgHHHE4PUVw0I5sIQt8hJq 5nH5dPqz4ITtCL9zjiJsExHuHKN3NZsAEQEAAcLAXwQYAQIACQUCU4xwFgIbDAAKCRCw3p3W KL8TL0P4B/9YWver5uD/y/m0KScK2f3Z3mXJhME23vGBbMNlfwbr+meDMrJZ950CuWWnQ+d+ Ahe0w1X7e3wuLVODzjcReQ/v7b4JD3wwHxe+88tgB9byc0NXzlPJWBaWV01yB2/uefVKryAf AHYEd0gCRhx7eESgNBe3+YqWAQawunMlycsqKa09dBDL1PFRosF708ic9346GLHRc6Vj5SRA UTHnQqLetIOXZm3a2eQ1gpQK9MmruO86Vo93p39bS1mqnLLspVrL4rhoyhsOyh0Hd28QCzpJ wKeHTd0MAWAirmewHXWPco8p1Wg+V+5xfZzuQY0f4tQxvOpXpt4gQ1817GQ5/Ed/wsDtBBgB CAAgFiEEhRJncuj2BJSl0Jf3sN6d1ii/Ey8FAlrd8NACGwIAgQkQsN6d1ii/Ey92IAQZFggA HRYhBFMtsHpB9jjzHji4HoBcYbtP2GO+BQJa3fDQAAoJEIBcYbtP2GO+TYsA/30H/0V6cr/W V+J/FCayg6uNtm3MJLo4rE+o4sdpjjsGAQCooqffpgA+luTT13YZNV62hAnCLKXH9n3+ZAgJ RtAyDWk1B/0SMDVs1wxufMkKC3Q/1D3BYIvBlrTVKdBYXPxngcRoqV2J77lscEvkLNUGsu/z W2pf7+P3mWWlrPMJdlbax00vevyBeqtqNKjHstHatgMZ2W0CFC4hJ3YEetuRBURYPiGzuJXU pAd7a7BdsqWC4o+GTm5tnGrCyD+4gfDSpkOT53S/GNO07YkPkm/8J4OBoFfgSaCnQ1izwgJQ jIpcG2fPCI2/hxf2oqXPYbKr1v4Z1wthmoyUgGN0LPTIm+B5vdY82wI5qe9uN6UOGyTH2B3p hRQUWqCwu2sqkI3LLbTdrnyDZaixT2T0f4tyF5Lfs+Ha8xVMhIyzNb1byDI5FKCb
  • Cc: hans@xxxxxxxxxxx, grub-devel@xxxxxxx, phcoder@xxxxxxxxx, xen-devel@xxxxxxxxxxxxx
  • Delivery-date: Tue, 20 Nov 2018 09:27:48 +0000
  • List-id: Xen developer discussion <xen-devel.lists.xenproject.org>
  • Openpgp: preference=signencrypt

On 09/11/2018 19:12, Daniel Kiper wrote:
> On Fri, Nov 02, 2018 at 01:37:29PM +0100, Juergen Gross wrote:
>> Add the needed code to setup the hypercall page for calling into the
>> Xen hypervisor.
>>
>> Signed-off-by: Juergen Gross <jgross@xxxxxxxx>
>> ---
>> V3: grub_xen_early_halt->grub_xen_panic (Roger Pau Monné)
>>     issue panic message (Roger Pau Monné)
>>     rewrite grub_xen_hypercall to avoid register variables (Daniel Kiper)
>> ---
>>  grub-core/kern/i386/xen/pvh.c | 69 
>> +++++++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 69 insertions(+)
>>
>> diff --git a/grub-core/kern/i386/xen/pvh.c b/grub-core/kern/i386/xen/pvh.c
>> index ac6181f4e..7e90a4538 100644
>> --- a/grub-core/kern/i386/xen/pvh.c
>> +++ b/grub-core/kern/i386/xen/pvh.c
>> @@ -20,14 +20,83 @@
>>  #include <grub/misc.h>
>>  #include <grub/memory.h>
>>  #include <grub/mm.h>
>> +#include <grub/i386/cpuid.h>
>> +#include <grub/i386/io.h>
>>  #include <grub/xen.h>
>>  #include <grub/machine/kernel.h>
>>
>> +#define XEN_CONSOLE_PORT   0xe9
> 
> I think that this is not PVH specific thing. Could you move this to
> more generic Xen header?
> 
>>  grub_uint64_t grub_rsdp_addr;
> 
> Hmmm... It seems to me that immediately after patch #8 GRUB2 build is
> broken. Is not it?

No. xen-pvh isn't yet selectable.

> 
>> +static char hypercall_page[GRUB_XEN_PAGE_SIZE]
>> +  __attribute__ ((aligned (GRUB_XEN_PAGE_SIZE)));
>> +
>> +static grub_uint32_t xen_cpuid_base;
>> +
>> +static void
>> +grub_xen_panic (const char *msg)
>> +{
>> +  const char *c;
>> +
>> +  for (c = msg; *c; c++)
>> +    grub_outb (*c, XEN_CONSOLE_PORT);
>> +
>> +  asm volatile ("hlt");
> 
> Should not you do something similar to grub-core/lib/i386/halt.c:stop() here?

Yes, as already replied to Roger.

> 
>> +}
>> +
>> +static void
>> +grub_xen_cpuid_base (void)
>> +{
>> +  grub_uint32_t base, eax, signature[3];
>> +
>> +  for (base = 0x40000000; base < 0x40010000; base += 0x100)
>> +    {
>> +      grub_cpuid (base, eax, signature[0], signature[1], signature[2]);
>> +      if (!grub_memcmp ("XenVMMXenVMM", signature, 12) && (eax - base) >= 2)
>> +    {
>> +      xen_cpuid_base = base;
>> +      return;
>> +    }
>> +    }
>> +
>> +  grub_xen_panic ("Found no Xen signature.\n");
> 
> "Found no Xen signature!\nSystem halted!\n"
> 
> Or maybe grub_xen_panic() should always add "System halted!\n".

Okay (the latter).

> 
>> +}
>> +
>> +static void
>> +grub_xen_setup_hypercall_page (void)
>> +{
>> +  grub_uint32_t msr, pfn, eax, ebx, ecx, edx;
>> +
>> +  grub_cpuid (xen_cpuid_base + 2, eax, ebx, ecx, edx);
>> +  msr = ebx;
>> +  pfn = (grub_uint32_t) (&hypercall_page[0]);
> 
> Could not you use hypercall_page alone here?

Hmm, yes.

> 
>> +
>> +  asm volatile ("wrmsr" : : "c" (msr), "a" (pfn), "d" (0) : "memory");
>> +}
>> +
>> +int
>> +grub_xen_hypercall (grub_uint32_t callno, grub_uint32_t a0,
>> +                grub_uint32_t a1, grub_uint32_t a2,
>> +                grub_uint32_t a3, grub_uint32_t a4,
>> +                grub_uint32_t a5 __attribute__ ((unused)))
>> +{
>> +  grub_uint32_t __res, __ign0, __ign1, __ign2, __ign3, __ign4;
>> +
>> +  asm volatile ("call *%[callno]"
>> +            : "=a" (__res), "=b" (__ign0), "=c" (__ign1), "=d" (__ign2),
>> +              "=S" (__ign3), "=D" (__ign4)
>> +            : "1" (a0), "2" (a1), "3" (a2), "4" (a3), "5" (a4),
> 
> I think that you can drop all __ign* variables if you specify proper
> registers in input argument. If this does not work you can use "+"
> modifier instead of "=" in the output argument.

I'll have a try.


Juergen


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.