Message ID | 20180927211322.16118-1-richard.henderson@linaro.org |
---|---|
Headers | show |
Series | target/arm: Rely on id regs instead of features | expand |
On 27 September 2018 at 22:13, Richard Henderson <richard.henderson@linaro.org> wrote: > There are more feature bits that could be converted, but I thought > I should show the work to this point to get feedback. > > This is the "v2" as compared to > http://lists.nongnu.org/archive/html/qemu-devel/2018-09/msg01849.html So having written the code both ways round, which do you think works out nicer? (or at least least-worst :-)) thanks -- PMM
On 02/10/2018 14:30, Peter Maydell wrote: > On 27 September 2018 at 22:13, Richard Henderson > <richard.henderson@linaro.org> wrote: >> There are more feature bits that could be converted, but I thought >> I should show the work to this point to get feedback. >> >> This is the "v2" as compared to >> http://lists.nongnu.org/archive/html/qemu-devel/2018-09/msg01849.html > > So having written the code both ways round, which do you think > works out nicer? (or at least least-worst :-)) Using the *FEATURE*() and FIELD_DP/EX() macros are a great improvement, the code is more concise, easier to review. I also appreciate the arm_features enum cleaned, using the previous macros on ISAR. I doubt this version works worst, but it definitively looks nicer, thus less bug-prone.
On 10/2/18 7:30 AM, Peter Maydell wrote: > On 27 September 2018 at 22:13, Richard Henderson > <richard.henderson@linaro.org> wrote: >> There are more feature bits that could be converted, but I thought >> I should show the work to this point to get feedback. >> >> This is the "v2" as compared to >> http://lists.nongnu.org/archive/html/qemu-devel/2018-09/msg01849.html > > So having written the code both ways round, which do you think > works out nicer? (or at least least-worst :-)) My preference is for v2. r~