This is the mail archive of the gdb-patches@sourceware.cygnus.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [eliz@is.elta.co.il: GDB: fine print in 387 info]



> What exactly is the additional information you want "info float" to
> show that "info all-registers" cannot show?

My understanding is that "info all-registers" is the way to display
the registers whose values the user can change.  Therefore, it _must_
show all the registers using *exactly* the names and the format that
users needs to work with when they change the register's value.

For example, since the FP registers are called st0, st1, etc., "info
all-registers" *must* use those names.  As another example, the
register that Bill called "fcs" includes, in addition to the code
selector, the last 11 bits of the opcode.  "info all-registers" must
present that register exactly as stored, because that's what the user
would have to use to change its value.

In contrast, "info float" is free from this limitation, and can
therefore present the information in more user-friendly fashion
without bothering about the original format of the data inside the
registers.

> "info float" is an opportunity to provide information that "info
> all-registers" cannot, but since the latter is allowed to print out
> nice human-readable versions of flags, etc. to the right of the raw
> value, it can provide quite a bit of information.

The problem is the real estate available for display.  There are only
so many rows and columns on any given screen, before you are forced to
go to small fonts and/or multi-page displays.  "info all-registers"
already takes up more than a screenful on many displays.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]