Equation: 120, when w=JUMPI, outcome of JUMPI depends on µs[1] too, so the exception Z and its coding

[efaysal]

In Exceptional Halting: Equation: 120,

(w ∈ {JUMP, JUMPI} ∧ µs[0] not in D(Ib)) --^^--

should be separated as the outcome of JUMPI is also depending on the value of µs[1]. Better to use the following:

(w =JUMP ∧ µs[0] not in D(Ib)) V (w =JUMPI ∧ µs[1] != 0 ∧ µs[0] not in D(Ib)) ... if µs[1] = 0, JUMPI is just shifting the pc by one, and it should be an opcode not data assuming that pc+1 is less than the total number of opdodes in Ib.

--^^-- comes to:

(w =JUMP ∧ µs[0] not in D(Ib)) V (w =JUMPI ∧ µs[1] != 0 ∧ µs[0] not in D(Ib)) V (w =JUMPI ∧ µs[1] = 0 ∧ pc+1 > Ic|), Paper.tex has my committed changes.

Also is the C++ coding of JUMPI used µs[1]?

Implicitly the condition related to µs[1] is tested in "if (m_stack[m_stack.size() - 2])", is this robust?

At least the topmost two elements from the stack should exist if not an issue is there because JUMPI needs, accordingly to the value of µs[1], to decide to jump to µs[0] or to just increment by one the current pc.

Much safe to test the existence of µs[1] && if (m_stack.size()>2) to decide.

case Instruction::JUMPI:
if (m_stack[m_stack.size() - 2]) \ only OK when m_stack.size() >2 ...
nextPC = verifyJumpDest(m_stack.back(), m_jumpDests);
m_stack.pop_back();
m_stack.pop_back();
break;

[gavofyork]

pc+1 > |c| is unnecessary; that case is already guaranteed valid by virtue of the explicit assertion that any value of code outside the specific code portion given (c) is defined as 0/STOP. Otherwise, this is sensible and should go in. Please reformulate accordingly and I'll merge.

[efaysal]

Done!

[gavofyork]

needs rebasing/merging

[efaysal]

I guess it's OK now:
git checkout master
Already on 'master'
Your branch is up-to-date with 'origin/master'.

[gavofyork]

still needs rebasing/merging

[axic]

@efaysal can you please rebase?

[efaysal]

After a number of trails, It shows this "This branch has no conflicts with the base branch
Only those with write access to this repository can merge pull requests." .

[efaysal]

I tried to fix the issues with Travis, I got this,

remote: Permission to ethereum/yellowpaper.git denied to efaysal.

[gavofyork]

this is now broken. please ensure it is merged properly. if you're unsure how to do this, please read up on how git works.

[efaysal]

I only need to reread a Pull Request Tutorial, indeed.

[gavofyork]

The identified issue needs correcting, but this PR still needs a proper resolution/rebase.

[efaysal]

I'm sharing with you the steps, everything was OK. However when I pushed as it was required,
I got this The requested URL returned error: 403. The logs of my steps are below,

faysal@faysal-VPCEH12FX:$ cd yellowpaper
faysal@faysal-VPCEH12FX:/yellowpaper$ git remote add upstream1 https://github.com/ethereum/yellowpaper
faysal@faysal-VPCEH12FX:/yellowpaper$ git remote -v
origin https://github.com/ethereum/yellowpaper (fetch)
origin https://github.com/ethereum/yellowpaper (push)
upstream https://github.com/ethereum/yellowpaper (fetch)
upstream https://github.com/ethereum/yellowpaper (push)
upstream1 https://github.com/ethereum/yellowpaper (fetch)
upstream1 https://github.com/ethereum/yellowpaper (push)
faysal@faysal-VPCEH12FX:/yellowpaper$ git fetch upstream1
From https://github.com/ethereum/yellowpaper

• [new branch] deploytest -> upstream1/deploytest
• [new branch] frontier -> upstream1/frontier
• [new branch] gh-pages -> upstream1/gh-pages
• [new branch] invalid_is_invalid -> upstream1/invalid_is_invalid
• [new branch] master -> upstream1/master
• [new branch] metropolis -> upstream1/metropolis
• [new branch] poc-4 -> upstream1/poc-4
• [new branch] superfluous -> upstream1/superfluous
  faysal@faysal-VPCEH12FX:/yellowpaper$ git reset --hard upstream1/superfluous
  HEAD is now at a3381de Various typos fixed thanks to 看K线图.
  faysal@faysal-VPCEH12FX:/yellowpaper$ git checkout master
  Previous HEAD position was a3381de... Various typos fixed thanks to 看K线图.
  Switched to branch 'master'
  Your branch is ahead of 'origin/master' by 1 commit.
  (use "git push" to publish your local commits)
  faysal@faysal-VPCEH12FX:/yellowpaper$ git add --all
  faysal@faysal-VPCEH12FX:/yellowpaper$ git commit -m "Initial commit-"
  On branch master
  Your branch is ahead of 'origin/master' by 1 commit.
  (use "git push" to publish your local commits)

You are currently rebasing.

nothing to commit, working directory clean
faysal@faysal-VPCEH12FX:~/yellowpaper$ git push -u origin master
Username for 'xxxxx'
Password for 'xxxxx':
remote: Permission to ethereum/yellowpaper.git denied to efaysal.
fatal: unable to access 'https://github.com/ethereum/yellowpaper/': The requested URL returned error: 403

[fulldecent]

This PR is getting all messed up.

Recommendation:

• Make a backup of your working directory
• Delete your fork on github
• Refork on github
• Redo the changes
• git push --force

[jamesray1]

I don't think this is changing anything compared to the latest (indeed it's outdated): \text{max}\!\left(\mindifficulty, {P(H)_{H}}_{\mathrm{d}} + \diffadjustment\times\homesteadmod + \expdiffsymb \right) & \text{otherwise}\\

[jamesray1]

Currently we have:

yellowpaper/Paper.tex

Lines 817 to 829 in fb260c1

	\begin{eqnarray}
	\boldsymbol{\sigma}' & \equiv & \begin{cases}
	\boldsymbol{\sigma} & \text{if} \quad \boldsymbol{\sigma}^{**} = \varnothing \\
	\boldsymbol{\sigma}^{**} & \text{otherwise}
	\end{cases} \\
	g' & \equiv & \begin{cases}
	0 & \text{if} \quad \boldsymbol{\sigma}^{**} = \varnothing \ \wedge \ \mathbf{o} = \varnothing \\
	g^{**} & \text{otherwise}
	\end{cases} \\ \nonumber
	z & \equiv & \begin{cases}
	0 & \text{if} \quad \boldsymbol{\sigma}^{**} = \varnothing \\
	1 & \text{otherwise}
	\end{cases} \\

I'm not sure what the benefit of including this is.

[jamesray1]

@efaysal, I put 919-920 into #616