> On 21 May 2015, at 21:35, Maciej W. Rozycki wrote:
>
> On Thu, 21 May 2015, Liviu Ionescu wrote:
>
>> p.s. and I think there are major differences between using -append to
>> pass arguments to the kernel and passing arguments to semihosting, this
>> being one of the reasons for suggesting a
On Thu, 21 May 2015, Liviu Ionescu wrote:
> p.s. and I think there are major differences between using -append to
> pass arguments to the kernel and passing arguments to semihosting, this
> being one of the reasons for suggesting a new command.
Out of curiosity, why do you think there are diff
> On 21 May 2015, at 19:56, Peter Maydell wrote:
>
> ... That's because your code is adding quotes, which I've already
> said I think is incorrect. If you took out the addition of
> quotes the two would be the same.
it might be incorrect for some obscure/broken guest implementations, but my
se
On 21 May 2015 at 17:36, Liviu Ionescu wrote:
>
>> On 21 May 2015, at 18:54, Peter Maydell wrote:
>>
>> On 21 May 2015 at 16:47, Liviu Ionescu wrote:
>>>
On 21 May 2015, at 18:29, Peter Maydell wrote:
There's no way to not offer the choice. In your syntax
this would be
> On 21 May 2015, at 18:54, Peter Maydell wrote:
>
> On 21 May 2015 at 16:47, Liviu Ionescu wrote:
>>
>>> On 21 May 2015, at 18:29, Peter Maydell wrote:
>>>
>>>
>>> There's no way to not offer the choice. In your syntax
>>> this would be
>>> --semihosting-cmdline "foo bar baz"
>>
>> not ex
On 21 May 2015 at 16:47, Liviu Ionescu wrote:
>
>> On 21 May 2015, at 18:29, Peter Maydell wrote:
>>
>>
>> There's no way to not offer the choice. In your syntax
>> this would be
>> --semihosting-cmdline "foo bar baz"
>
> not exactly. in my implementation this would arrive in the guest
> code as
> On 21 May 2015, at 18:29, Peter Maydell wrote:
>
>
> There's no way to not offer the choice. In your syntax
> this would be
> --semihosting-cmdline "foo bar baz"
not exactly. in my implementation this would arrive in the guest code as
argc=1, argv[0]="foo bar baz", which will be understood
On 21 May 2015 at 16:24, Liviu Ionescu wrote:
>
>> On 21 May 2015, at 18:06, Peter Maydell wrote:
>>
>> ...
>> That's their choice... It is still an array of substrings,
>> but if the user only provides one string then we have an
>> array of length 1.
>
> this was my initial point, as long as you
> On 21 May 2015, at 18:06, Peter Maydell wrote:
>
> ...
> That's their choice... It is still an array of substrings,
> but if the user only provides one string then we have an
> array of length 1.
this was my initial point, as long as you offer the choice, you no longer have
consistent usage
On 21 May 2015 at 15:57, Liviu Ionescu wrote:
>
>> On 21 May 2015, at 17:33, Peter Maydell wrote:
>>
>> On 21 May 2015 at 15:28, Liviu Ionescu wrote:
>>> from what I understood, for MIPS the emulator must be called with a
>>> sequence of ,arg=xxx, while for ARM it needs to be called with a singl
> On 21 May 2015, at 17:33, Peter Maydell wrote:
>
> On 21 May 2015 at 15:28, Liviu Ionescu wrote:
>> from what I understood, for MIPS the emulator must be called with a
>> sequence of ,arg=xxx, while for ARM it needs to be called with a single
>> long string including the entire command line.
On 21 May 2015 at 15:28, Liviu Ionescu wrote:
> from what I understood, for MIPS the emulator must be called with a
> sequence of ,arg=xxx, while for ARM it needs to be called with a single
> long string including the entire command line.
You can use a sequence of ,arg=xxx on ARM too. It's just t
> On 21 May 2015, at 16:57, Leon Alrae wrote:
>
> ... pseudo-bootloaders in MALTA and Fulong2e have hardcoded argc to 2
> (-kernel and -append).
I don't know the specifics of other emulated platforms, but to me the whole
issue looks messy, and the details of using the newly added arg option j
On 21/05/2015 15:01, Peter Maydell wrote:
> Well, it will presumably mean that on a MIPS semihosting case
> if the user tries -append "some command line" it won't do what
> they expect...
Yes, but I think the user should use semihosting-config arg, which is
more natural in semihosting context, rat
On 21 May 2015 at 14:57, Leon Alrae wrote:
> I've been considering that, and I think we are better off without
> splitting -append in semihosting mode because this is at least
> consistent with MIPS machines in QEMU (in non-semihosting mode) --
> pseudo-bootloaders in MALTA and Fulong2e have hardc
On 20/05/2015 09:54, Peter Maydell wrote:
> On 20 May 2015 at 09:11, Leon Alrae wrote:
>> Agreed. We could initialize semihosting.argv[0] with -kernel and argv[1]
>> with -append string if in semihosting mode and no semihosting args have
>> been specified. I'll update it in v4.
>
> I'm tempted to
> On 20 May 2015, at 18:11, Peter Maydell wrote:
>
> --semihosting-config,arg='program --opt1 "path 1" --opt2 "path 2"'
ok, thank you, this seems manageable.
Liviu
On 20 May 2015 at 15:59, Liviu Ionescu wrote:
>
>> On 20 May 2015, at 17:40, Peter Maydell wrote:
>>
>> ... The ARM semihosting API doesn't mandate any kind of quoting, and we
>> can't know what the guest application expects,
>
> ok (btw, my embedded semihosting code is happy with both quotes or
> On 20 May 2015, at 17:40, Peter Maydell wrote:
>
> ... The ARM semihosting API doesn't mandate any kind of quoting, and we
> can't know what the guest application expects,
ok (btw, my embedded semihosting code is happy with both quotes or apostrophes)
> so the user is going
> to have to sup
On 20 May 2015 at 15:31, Liviu Ionescu wrote:
>
>> On 20 May 2015, at 17:18, Peter Maydell wrote:
>>
>> PS: the glib function g_strjoinv("", null_terminated_string_array)
>> is probably better than reimplementing the string-concatenation
>> wheel.
>
> I'm not familiar with g_strjoinv(), but the a
> On 20 May 2015, at 17:18, Peter Maydell wrote:
>
> PS: the glib function g_strjoinv("", null_terminated_string_array)
> is probably better than reimplementing the string-concatenation
> wheel.
I'm not familiar with g_strjoinv(), but the args concatenate logic should
protect args containing s
On 20 May 2015 at 14:49, Liviu Ionescu wrote:
> personally I would add a pointer (cmdline?) in the semihosting
> structure and a getter (semihosting_get_cmdline()?). at first call,
> the getter would concatenate all args and store the result in the
> structure (the concatenate function is there),
> On 20 May 2015, at 12:31, Leon Alrae wrote:
>
> ... I’ll update arm-semi.c as well to concatenate all args.
you can use parts of my initial patch for the arm semihosting related changes.
---
personally I would add a pointer (cmdline?) in the semihosting structure and a
getter (semihosting_
On 20/05/2015 12:12, Liviu Ionescu wrote:
> yes, that's correct, my comment was related to the above quotes from your
> message, in the first part you mention "the path won't be passed to
> semihosting", which is not consistent with the actual behaviour, since the
> entire kernel path is passed
> On 20 May 2015, at 11:51, Leon Alrae wrote:
>
> On 20/05/2015 09:30, Liviu Ionescu wrote:
>>
>>> On 20 May 2015, at 11:11, Leon Alrae wrote:
>>>
>>> like "If both are specified, -kernel/-append are ignored (-kernel is
>>> used to load an image, but the path won't be passed to semihosting)"
On 20/05/2015 09:54, Peter Maydell wrote:
> Well, with these patches alone we end up with a documented but broken
> option. For reasons of consistency (again) I don't want the new
> option in the tree without it being supported by ARM, even if the
> MIPS code that uses it goes in.
Fair enough. I’l
On 20 May 2015 at 09:11, Leon Alrae wrote:
> On 18/05/2015 17:18, Peter Maydell wrote:
>> You need to say how this interacts with the -kernel/-append option
>> (ie what happens if you specify both).
>
> I don't see any correlation between semihosting options and "-append"
> which is described as "
On 20/05/2015 09:30, Liviu Ionescu wrote:
>
>> On 20 May 2015, at 11:11, Leon Alrae wrote:
>>
>> like "If both are specified, -kernel/-append are ignored (-kernel is
>> used to load an image, but the path won't be passed to semihosting)"
> ...
>> We could initialize semihosting.argv[0] with -ker
> On 20 May 2015, at 11:11, Leon Alrae wrote:
>
> like "If both are specified, -kernel/-append are ignored (-kernel is
> used to load an image, but the path won't be passed to semihosting)"
...
> We could initialize semihosting.argv[0] with -kernel
here you have a small contradiction, both in
On 18/05/2015 17:18, Peter Maydell wrote:
> On 8 May 2015 at 12:41, Leon Alrae wrote:
>> Add new "arg" sub-argument to the --semihosting-config allowing to pass
>> multiple input argument separately. It is required for example by UHI
>> semihosting to construct argc and argv.
>>
>> Signed-off-by:
On 8 May 2015 at 12:41, Leon Alrae wrote:
> Add new "arg" sub-argument to the --semihosting-config allowing to pass
> multiple input argument separately. It is required for example by UHI
> semihosting to construct argc and argv.
>
> Signed-off-by: Leon Alrae
> ---
> include/exec/semihost.h | 12
Add new "arg" sub-argument to the --semihosting-config allowing to pass
multiple input argument separately. It is required for example by UHI
semihosting to construct argc and argv.
Signed-off-by: Leon Alrae
---
include/exec/semihost.h | 12
qemu-options.hx | 19
32 matches
Mail list logo